Preliminary reporting patterns of suicide ideation and attempt among Native American adolescents in two samples

Abstract

Objective:

Native American adolescents are disproportionately burdened by suicidality. Here we examine patterns of reporting of suicide ideation and suicide attempt among Native American youth compared those from other ethnoracial backgrounds, as this data is important for grounding commonly subscribed to frameworks of suicide risk (e.g., ideation-to-action).

Method:

Data are from the Youth Risk Behavior Surveillance Survey (N=54,243; grades 9–12; 51.0% female) and Minnesota Student Survey (N=335,151; grades 8, 9, 11; 50.7% female). Comparing Native American youth to peers from other ethnoracial backgrounds, we examined two suicide reporting patterns: 1) odds of reporting suicide attempt among those who reported ideation and 2) odds of reporting suicide ideation among those who reported an attempt.

Results:

Across both samples, when reporting suicide ideation, youth from other ethnoracial backgrounds were 20-55% less likely than Native American youth to also report attempt. While few consistent differences were observed between Native American youth and those from other racial minority backgrounds in patterns of co-reporting suicide ideation and attempt across samples, White youth were between 37-63% less likely than Native American youth to report a suicide attempt without also reporting ideation.

Conclusions:

The increased odds of engaging in a suicide attempt with or without reporting ideation question the generalizability of widely-held frameworks of suicide risk to Native American youth and have important implications for suicide risk monitoring. Future research is needed to illuminate how these behaviors unfold over time and the potential mechanisms of risk for engaging in suicide attempts in this disproportionately burdened group.

Suicidal behaviors are a leading cause of death for young people in the United States (Centers for Disease Control and Prevention (CDC), National Center for Health Statistics, 2020). Acknowledging that Native American¹ peoples are highly diverse and there are differences in suicide mortality rates between Native communities (Beals, Manson, Mitchell, Spicer, & Beals, 2003), Native American youth on the whole are more likely than youth from all other ethnoracial backgrounds to die by suicide (CDC, NCHS, 2020; Substance Abuse and Mental Health Services Administration, 2017). Specifically, in 2020, non-Hispanic Native American youth ages 10 to 19-years-old died by suicide at a rate of 22.03 per 100,000, with the next highest demographic being non-Hispanic White² youth at 7.41 deaths per 100,000 (CDC, 2020). Research has also consistently identified a disproportionate burden of non-fatal suicidal thoughts and behaviors among Native American youth as compared to their age-mates (Bai & Specker, 2019; Campbell & Eggerling-Boeck, 2006; Qiao & Bell, 2017). Suicide in Native American populations has been noted as an important area for future research by the CDC and the Action Alliance for Suicide Prevention (U.S. Department of Health and Human Services, 2010). Despite this endorsement, this literature remains remarkably scant (Olson & Wahab, 2006; Rey et al., 2022) and efforts to identify effective suicide interventions for Native peoples have had limited success (Pham et al., 2021). Moreover, the rates of suicide among Native peoples continue to remain stable or rise (CDC, 2021). In order to promote effective suicide prevention for Native American young people, it is critically important to examine whether widely-held frameworks about suicide, which likely permeate clinical practice and the public consciousness, hold true for Native American youth.

A pattern of increased suicide risk among Native young people occurs in the shared context of systemic racism and oppression that traces back to colonization and contributes to ongoing historical trauma (e.g., collective and compounding emotional and psychological wounding across generations; Brave Heart, 1998; Brave Heart et al., 2011). It is this historical context that also places Native American youth at higher risk of experiencing a host of socioecological stressors, such as poverty and maltreatment (e.g., Bohn, 2003; Brockie et al., 2015; Malcoe et al., 2004). Within this context of varying stressors, a range of critical risk factors, such as impulsivity and acquired capability for suicide, may take hold (e.g., Berger & Sarnyai, 2015; Brooks et al., 2020; Simpson et al., 2012). While some facets of Native experiences are shared with other minority groups in the U.S.– such as minority stress (Meyer, 2003), racism, prejudice– others are unique– for example a history of forced relocation by the United States government (Biolsi, 2005) into regions with little economic development (Davis, Roscigno, & Wilson, 2016), the forced removal of Native American children from their families and homes into boarding schools (Trafzer, Keller, & Sisquoc, 2006), and the criminalization of Native American spiritual practices in the United States until 1978 (Native American Rights Fund, 1979). Together, these factors are thought to impact suicide risk trajectories for Native American youth in a way that might differ from youth of other ethnoracial backgrounds (e.g., Burnette & Figley, 2016; O’Keefe et al., 2022).

The suicide literature has seen the recent development of an “ideation-to-action” suicide framework which posits that the development of suicide ideation and the progression from suicide ideation to suicide behaviors are two distinct processes that are explained by unique factors or mechanisms (Klonsky & May, 2015). The ideation-to-action framework explicitly states that suicide thoughts are a “prerequisite” to suicide behavior and more broadly assumes that there is a common progression where individuals move along a continuum of suicide risk that involves increasing severity of suicide ideation, intent, planning, and attempts (Klonsky & May, 2014). This framework highlights the importance of differentiating those who think about suicide from those who act on thoughts of suicide in hopes of identifying intervention points to prevent suicidal behavior from occurring (Klonsky & May, 2014; May & Klonsky, 2016).

There are two ways in which the ideation-to-action framework may fall short when considering its application to Native American young people. First, this research has been conducted with samples that are primarily White (Huang, Ribeiro, & Franklin, 2020; Mars et al., 2019; May & Klonsky, 2016), with little consideration of differences between ethnoracial or cultural groups, thus the assumptions of this framework may not hold true for ethnoracially diverse youth, including Native American youth (Chu, Goldblum, Floyd, & Bongar, 2010). Second, there are a number of people who attempt and die by suicide who do not report having, or are not recorded to have, suicide ideation prior to their attempt (Bolton et al., 2014; Jordan & McNiel, 2020; LeMaster, Beals, Novins, & Manson, 2004; McKean, Pabbati, Geske, & Bostwick, 2018). Reporting suicide attempts without reporting prior or concurrent suicide ideation violates the ideation-to-action framework and is a pattern that is starting to be identified among Native populations. In one study, a greater percentage of Northern Plains adolescents and adults reported engaging in a past suicide attempt than reported experiencing past suicide ideation or planning (LeMaster et al., 2004). Further, the largest epidemiological sample of Native American reservation populations to date, which included two closely related Northern Plains tribes and a Southwestern tribe (Beals et al., 2005), found that those who reported engaging in a suicide attempt were less likely than the U.S. general population to endorse suicidal thoughts or plans (Bolton et al., 2014). Findings across both studies suggest that in Native American populations, suicide ideation may not necessarily precede suicide attempt as is generally assumed in the ideation-to-action framework. However, this existing work does not disaggregate Native American youth from adults (e.g., samples ages 15-54 years old), which may be particularly problematic given that suicide peaks in early adulthood for Native populations (CDC, 2020) and impulsive traits (thus, possibly impulsive attempts) are more common among young people who die by suicide as compared to older individuals (McGirr et al., 2008). Indeed, as highlighted in a sample of Apache youth, most Native American youth who have made a non-fatal suicide attempt report that their attempts were impulsive in nature (Cwik et al., 2015).

It is an important step for suicide prevention to examine features of the ideation-to-action framework among Native American youth because frameworks formulated from research on primarily White populations are not necessarily likely to seamlessly apply to Native American youth. That is, although researchers have long emphasized the importance of culturally sound theoretical frameworks and the incorporation of Indigenous world-views (e.g., Chu et al., 2010; Gone, 2007; O’Keefe et al., 2018), clinical professionals and the broader community likely make assumptions about suicide risk progression that align with the ideation-to-action framework (e.g., Wyder et al., 2021), which may be fundamentally misaligned with the phenomenon underlying Native suicide. And, in instances where risk trajectories violate the ideation-to-action framework (e.g., attempt without ideation), this misalignment could present a gap in suicide prevention approaches that leaves a subset of Native American youth particularly vulnerable (Wiglesworth, 2022). Thus, providing evidence of a misalignment by examining patterns of suicide risk in adolescence is a crucial step in moving theories of suicide among Native young people forward.

In this study, we examine patterns of reporting of suicide ideation and attempt among Native American adolescents as compared to their age-mates from other ethnoracial backgrounds in two separate large-scale samples with the goal of better understanding features of suicide risk reporting in this at-risk demographic. A recent systematic review noted that only three of the 45 studies (two dissertation studies, one peer-reviewed publication) examined suicide ideation as a risk factor for suicide attempt in Native American youth (Fetter et al., 2022). While these studies show that suicide ideation is related to suicide attempts among Native American youth (Howard-Pitney, LaFromboise, Basil, September, & Johnson, 1992), they do not delineate how commonly those who report ideation also report attempts (e.g., the proportion of youth who show a possible progression of risk that is consistent with the ideation-to-action framework), whether a subgroup of those who reported attempts do not report ideation (e.g., violating the ideation-to-action framework), or how these patterns of suicide risk reported by Native American youth are (dis)similar to those reported by their peers from other ethnoracial backgrounds (e.g., evidencing potential cultural differences).

To address these gaps, we will examine three reporting patterns: For descriptive purposes, we will examine the rates of reporting for suicide ideation and attempt in the past year. It is well-documented in the literature that suicide ideation and suicide attempt are elevated among Native American youth, but nevertheless, documenting this pattern in these samples will undergird our study aims. In Aim 1 of our study, among those who endorse suicide ideation, we will examine if there are ethnoracial differences in the likelihood of also reporting suicide attempt. Given the paucity of past research, we do not have a priori hypotheses. In Aim 2, among those who report suicide attempt, we will examine whether Native American youth are less likely than their peers to endorse suicide ideation. We predict that reports of suicide attempt will occur outside of the context of reports of suicide ideation more frequently for Native adolescents when compared to youth from other ethnoracial backgrounds. If our results yield ethnoracial differences, we will have garnered preliminary evidence that the patterns and corresponding assumptions of the ideation-to-action framework of the general population studies are not directly generalizable to Native American youth and illustrated the need for additional research that focuses on prospectively characterizing patterns of risk progression among Native young people.

Method

Data are from two epidemiologically informed screening surveys that use single items to identify constructs of interest across varying time scales. These surveys demonstrate minor modifications across years; however, the questions largely represent similar constructs and can be combined with careful consideration. Descriptions of survey items and coding are in Table 1. Data were de-identified prior to researcher access and we received an IRB exemption from the University of Minnesota for use of the Minnesota Student Survey. The authors report there are no competing interests to declare regarding this work.

Table 1.

Survey measures, coding, and recoding

Construct	Survey	Question	Response Options (Coding)
Sex	YRBSS	What is your sex?	Male (0), Female (1)
	MSS	What is your biological sex?	Male (0), Female (1)
Age	YRBSS	How old are you?	≤12, 13, 14, 15, 16, 17, ≥18 years old (ordinal)
	MSS	How old are you?	≤9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, ≥21 years old (ordinal)
Race/Ethnicity	YRBSS	2001: How do you describe yourself? 2005-2017: What is your race? (Select one or more responses) 2005-2017: Are you Hispanic or Latino?	2001: Same as 2005-2017, plus Hispanic/Latino, Multiracial - Hispanic/Latino, Multiracial - non-Hispanic/Latino 2005-2017: American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Other Pacific Islander, White 2005-2017: Yes (1), No (0) *Coded to indicate monoracial of one category or multiracial
	MSS	How do you describe yourself? (Mark all that apply)	American Indian or Alaska Native, Asian or Asian American, Black, African or African American, Hispanic or Latino/Latina, Native Hawaiian or Other Pacific Islander, White *Recoded to indicate monoracial of one category or multiracial
Suicide ideation	YRBSS	During the past 12 months, did you ever seriously consider attempting suicide?	Yes (1), No (0)
	MSS	Have you ever seriously considered attempting suicide?	No (0), Yes during the last year (1), Yes- more than a year ago (2) *Recoded to Yes in the past year (1), No/Not in the past year (0)
Suicide attempt	YRBSS	During the past 12 months, how many times did you actually attempt suicide?	0 times, 1 time, 2 or 3 times, 4 or 5 times, 6 or more times *Recoded into Yes, No
	MSS	Have you ever actually attempted suicide?	No (0), Yes during the last year (1), Yes- more than a year ago (2) *Recoded to Yes in the past year (1), No/Not in the past year (0)
Depressive symptoms	YRBSS	During the past 12 months, did you ever feel so sad or hopeless almost every day for two weeks or more in a row that you stopped doing some usual activities?	Yes (1), No (0)
	MSS	2013: During the last 12 months, have you had SIGNIFICANT problems with: Feeling very trapped, lonely, sad, blue, depressed or hopeless about the future? 2016-2019: Over the last 2 weeks, how often have you been bothered by: Little interest or pleasure in doing things? Feeling down, depressed or hopeless?	2013: Yes (1), No (0) 2016-2019: Not at all (0), Several Days (1), More than half the days (2), Nearly every day (3) *Recoded to Yes (1), No (0) for consistency across years

Note: When Minnesota Student Survey questions differ across years, all items are presented along with their corresponding dates.

Sample 1: Youth Risk Behavior Surveillance Survey

Participants for this study were drawn from the 2001, 2005, 2009, 2013, 2017 years of the Youth Risk Behavior Surveillance Survey (YRBSS), which is administered to youth in grades 9–12. While the YRBSS is administered bi-annually, YRBSS timepoints were chosen four years apart to limit the possibility of participant overlap. The final analytic sample consisted of 54,243 adolescents approximately 12–18 years old. The weighted race/ethnicity breakdown of the sample was 0.7% Native American, 3.5% Asian/ Asian American, 12.4% Black/ African American, 10.0% Hispanic/Latinx, 11.4% Multiracial, 0.8% Native Hawaiian/Pacific Islander, and 61.2% White/ European American (Table 2 for full demographics). The YRBSS is conducted by the CDC and designed to draw a nationally- representative sample of high school aged youth. The YRBSS administers self-report surveys in participating high schools across the country; state, Territory, and Tribal participation in the survey has varied over time (CDC, 2020). While specific Tribal communities included in the YRBSS sampling were able to decide whether data collected from their schools was released publicly, Native American youth were surveyed at Tribal and non-Tribal schools. This variability in sampling, and the fact that the YRBSS does not collect information on participant Tribal affiliation, means it is not possible to determine which Tribes and cultural groups are represented in the Native American sample for any given year.

Table 2.

Demographics and clinical factors

Factor, n (%)a	Native American	Asian/ Asian American	Black/ African American	Hispanic/ Latinx	Multiracial	Native Hawaiian/ Pacific Islander	White/ European American
Youth Risk Behavior Survey, N	556	2,054	9,949	8,644	7,302	483	25,255
Age, M (SE)	16.20 (1.16)	16.36 (1.07)	16.43 (1.06)	16.37 (1.06)	16.32 (1.10)	16.41 (1.09)	16.37 (1.05)
Sex
Male	316 (52.0%)	994 (50.1%)	4,693 (48.3%)	4,140 (48.9%)	3,517 (49.4%)	252 (47.4%)	12,684 (50.3%)
Female	240 (48.0%)	1,060 (49.9%)	5,256 (51.7%)	4,504 (51.1%)	3,785 (50.6%)	231 (52.6%)	12,571 (49.7%)
Depressive Symptoms	181 (32.8%)	577 (28.6%)	2,937 (29.0%)	2,898 (34.1%)	2,652 (36.0%)	163 (33.1%)	6,832 (26.9%)
Suicide ideation	128 (22.8%)	309 (15.8%)	1,369 (13.7%)	1,312 (15.1%)	1,521 (21.1%)	95 (20.2%)	4,160 (16.4%)
Suicide attempt	83 (14.0%)	131 (6.7%)	822 (8.3%)	784 (8.8%)	833 (10.9%)	57 (12.5%)	1,639 (6.4%)
Suicide ideation + attempt	64 (49.3%)	109 (34.3%)	675 (48.9%)	699 (50.7%)	745 (46.7%)	50 (53.6%)	1,515 (36.1%)
Suicide attempt − ideation	19 (19.9%)	22 (18.8%)	147 (19.0%)	85 (13.1%)	88 (9.4%)	7 (13.3%)	124 (6.8%)
Minnesota Student Survey, N	3,917	19,456	18,193	12,390	37,462	599	243,134
Age, M (sd)	14.70 (1.26)	14.80 (1.35)	14.71 (1.33)	14.79 (1.35)	14.69 (1.28)	14.89 (1.38)	14.87 (1.32)
Sex
Male	2,170 (55.4%)	9,548 (49.1%)	9,064 (49.8%)	5,991 (48.4%)	17,894 (47.8%)	354 (59.1%)	119,892 (49.3%)
Female	1,740 (44.6%)	9,882 (50.9%)	9,105 (50.2%)	6,389 (51.6%)	19,509 (52.2%)	243 (40.6%)	123,032 (50.7%)
Depressive Symptoms	1,162 (29.7%)	4,518 (23.2%)	3,712 (20.4%)	3,055 (24.7%)	10,558 (28.2%)	167 (27.9%)	46,489 (19.1%)
Suicide ideation	733 (18.7%)	2,279 (11.7%)	1,889 (10.4%)	1,640 (13.2%)	6,572 (17.5%)	86 (14.4%)	26,716 (11.0%)
Suicide attempt	340 (8.7%)	614 (3.2%)	683 (3.8%)	646 (5.2%)	2,526 (6.7%)	47 (7.8%)	7,246 (3.0%)
Suicide ideation + attempt	307 (41.9%)	545 (23.9%)	550 (29.1%)	588 (35.9%)	2,349 (35.7%)	40 (46.5%)	6,793 (25.4%)
Suicide attempt − ideation	33 (9.7%)	69 (9.6%)	133 (19.5%)	58 (9.0%)	177 (7.0%)	7 (14.9%)	453 (6.3%)

Note:

^aPercentages provided for the Youth Risk Behavior Survey are weighted percentages based on complex sampling design and may not reflect the non-weighted n’s provided. Suicide ideation + attempt = the number and percentage of those who endorse suicide ideation that also endorse an attempt; Suicide attempt − ideation = the number and percentage of those who endorse a suicide attempt that do not endorse suicide ideation.

Sample 2: Minnesota Student Survey

Participants for this study come from the 2013, 2016, and 2019 Minnesota Student Survey (MSS) which is administered to youth in grades 8, 9, and 11. To limit the possibility of participant overlap, 8^th graders from the 2013 and 2016 surveys were excluded from the sample. Our final analytic sample of 335,151 adolescents approximately 12–19 years old. The race/ethnicity breakdown of the sample was 1.2% Native American, 5.8% Asian/ Asian American, 5.4% Black/ African American, 3.7% Hispanic/Latinx, 11.2% Multiracial, 0.2% Native Hawaiian/Pacific Islander, and 72.5% White/ European American (Table 2 for full demographics). The MSS is a collaboration between the Minnesota Departments of Education, Health, Human Services, and Public Safety and is an anonymous statewide school-based survey administered every three years. All public, nonpublic, charter, and Tribal schools in the state of Minnesota are invited to participate. The MSS is recorded to capture more than half of the 8th, 9th, and 11th grade students in the state of Minnesota (‘Minnesota Student Survey’, 2021). Native American youth included in the MSS are primarily Anishinaabe/Ojibwe and Dakota/Lakota, though about a third report an “other” tribal affiliation.

Measures

Measures across both studies, as well as notes about their coding are in Table 1. Written description of the measures can be found in the Supplemental Method.

Data Analysis Plan

The YRBSS utilizes a multi-stage cluster sampling design to obtain a nationally representative sample of high school students (CDC, 2017). When using YRBSS data, it is essential to account for the complex sampling design, which weights data to adjust for student response rates based on students’ grade, sex, and race and unequal selection probabilities (Bell et al., 2012). The current project followed CDC guidelines for data analyses of the YRBSS through the use of the ‘survey’ package in R (Lumley, 2021). The complex sampling analysis plan accounts for the sample weights, stratification, and selection probabilities; details can be found in the YRBSS Data Users’ Guide (CDC, 2017). Weighed sample sizes are reported in Table S1.

The following analytic plan was executed for both studies. Analyses were conducted in R version 3.6.2. First, descriptive analyses were conducted separately across racial groups to examine group differences in sex, age, depressive symptoms, and rates of suicide ideation and attempt. We used multivariable logistic regression models to address each of our three study aims, controlling for age, sex, and depressive symptoms and included race as the primary predictor (with Native American youth as the reference group). Survey year was considered as an additional covariate but given that results remained largely similar with and without this variable, survey year was not retained. Adjusted odds ratios are reported and effect sizes are interpreted according to guidance provided by Chen and colleagues (2010) for epidemiological studies (for positive associations [e.g., OR ≥ 1], cut offs are 1.68 for small, 3.47 for medium, 6.71 for large effects, whereas for negative associations [e.g., OR < 1], cut offs are 0.60 for small, 0.30 for medium, and 0.15 for large effects).

Results

Group differences in age and sex were identified between racial groups, most notably with Native American youth being more likely than their peers to be younger (with the exception of Multiracial youth in YRBSS and Black/African American and Multiracial youth in MSS) and male (with the exception of Native Hawaiian/Pacific Islander youth across samples; Table S2, S5). Native American adolescents were more likely than Asian/Asian American, Black/African American, Hispanic/Latinx, and White/European American youth to report suicide ideation and attempt in the past year across both samples (p’s<0.01; Table S3, S6). Multiracial youth reported lower rates of suicide attempt as compared to Native American youth but did not differ significantly in rates of suicide ideation. Finally, Native Hawaiian/Pacific Islander youth were less likely to report suicide ideation compared to Native American youth in the MSS, but otherwise did not show significant differences in reports.

Aim 1: Suicide attempt patterns among those who reported suicide ideation

Youth Risk Behavior Surveillance Survey.

In the YRBSS, 8,894 youth reported a past year suicide ideation, of whom 43% also reported a past year suicide attempt. In this sample, those who reported suicide ideation and attempt demonstrated lower odds of being male (AOR [95% CI] = 0.85 [0.76-0.94]), and higher odds of being younger (AOR [95% CI] = 0.88 [0.76-0.94]) and reporting depressive symptoms (AOR [95% CI] = 2.63 [2.07-2.67]) than those who only reported suicide ideation (p’s<0.01; Table S2). When compared to Native American youth, Asian/Asian American youth (AOR [95% CI] = 0.55 [0.33-0.92], p=0.022) and White/European American youth (AOR [95% CI] = 0.60 [0.40-0.90], p=0.013) who reported suicide ideation showed significantly lower odds of also reporting a suicide attempt (Figure 1a). The reports of Black/African American, Hispanic/Latinx, Multiracial, and Native Hawaiian/Pacific Islander youth did not differ significantly from those of Native American youth (Table S4).

Figure 1.

Plots indicating the odds (and 95% confidence intervals) of reporting each outcome variable for youth from each separate ethnoracial background as compared to Native American youth (set as the reference group).

Minnesota Student Survey.

In the MSS, 39,915 youth reported a past year suicide ideation, of whom 28.0% reported a past year suicide attempt. Compared to those who reported only suicide ideation, those who reported both suicide ideation and attempt had lower odds of being male (AOR [95% CI] = 0.82 [0.78-0.86), and higher odds of being younger (AOR [95% CI] = 0.96 [0.94-0.98]) and reporting past 12-month depressive symptoms (AOR [95% CI] = 1.54 [1.47-1.62]; p’s < 0.001). Results revealed that, with the exception of Native Hawaiian/Pacific Islander youth (AOR [95% CI] = 1.23 [0.78-1.93], p=0.377) youth from all other racial or ethnic backgrounds who reported suicide ideation had significantly lower odds of reporting engaging in a suicide attempt compared to Native American youth (Table S7). Specifically, medium effects were found for comparisons of White/European American (AOR [95% CI] = 0.49 [0.42-0.57], p<0.001) and Asian/Asian American (AOR [95% CI] = 0.45 [0.38-0.54], p<0.001) youth, and small effects were found for comparisons of Black/African American (AOR [95% CI] = 0.61 [0.51-0.73], p<0.001), Hispanic/Latinx (AOR [95% CI] = 0.80 [0.67-0.96], p=0.017), and Multiracial (AOR [95% CI] = 0.78 [0.66-0.91], p=0.001) youth (Figure 1).

Aim 2: Suicide ideation patterns among those who reported suicide attempt

Youth Risk Behavior Surveillance Survey.

In the YRBSS, 4,349 youth reported a past year suicide attempt with 11.3% of those youth not reporting concurrent suicide ideation. Those who reported suicide attempt without suicide ideation had significantly higher odds of being male (AOR [95% CI] = 2.41 [1.88-3.10]) and lower odds of reporting depressive symptoms (AOR [95% CI] = 0.19[0.14-0.26]) as compared to those who reported both suicide attempt and suicide ideation (p’s<0.001). White/European American youth (AOR [95% CI] = 0.32 [0.17-0.60], p<0.001) had significantly lower odds of reporting suicide attempt without also reporting suicide ideation as compared to Native American youth, with the difference in odds indicating a medium effect size (Figure 1a). However, significant differences were not observed between Native American youth and youth from other minoritized ethnoracial backgrounds (Table S4).

Minnesota Student Survey.

In the MSS, 12,102 youth reported a past year suicide attempt with 7.7% not reporting concurrent suicide ideation. Similar to the YRBSS, in the MSS, those who reported suicide attempt without suicide ideation were more likely to be male (AOR [95% CI] = 2.40 [2.08-2.75]) and older (AOR [95% CI] =1.11 [1.05-1.17]) and less likely to report depressive symptoms (AOR [95% CI] = 0.38 [0.33-0.44]) as compared to those who reported both suicide attempt and suicide ideation (p’s<0.001). As compared to Native American youth, White/European American youth (AOR [95% CI] = 0.63 [0.44-0.94], p=0.019) had significantly lower odds of reporting a suicide attempt without also reporting suicide ideation (Figure 1b). Conversely, Black/African American youth had significantly higher odds than Native American youth of reporting a suicide attempt without reporting ideation (AOR [95% CI] = 1.97 [1.31-3.03], p=0.002). No significant differences were observed for Asian/Asian American, Hispanic/Latinx, Multiracial, and Native Hawaiian/Pacific Islander youth versus Native American youth (Table S7).

Discussion

The current project aimed to address a critical gap within the literature on suicide risk by examining concurrent patterns of reporting of suicide ideation and attempt among Native American youth and their peers from other ethnoracial backgrounds in two relatively large, epidemiologically informed samples. Providing foundational evidence for investigating the primary study aims, Native American youth more frequently reported suicidal thoughts and behavior compared to those in most other ethnoracial groups (Bryan & Bryan, 2015; Manzo et al., 2015; YRBSS, 2020). Additionally, youth from other ethnoracial groups who reporting thinking about suicide less frequently reported suicide attempts than did Native American youth, and this pattern was robust across samples when comparing White/European American and Asian/Asian American youth with Native American youth. Finally, a subsample of youth demonstrating a reporting pattern of endorsing engage in a suicide attempt without concurrent ideation. Importantly, this response pattern was reported by fewer White/European American youth than Native American youth and more Black/African American youth than Native American youth. Documenting these patterns highlights the importance of continuing to strive to understand how suicide risk progresses among Native American youth.

Ideation-to-Action and Suicide Risk Assessment

We examined patterns of co-reporting of suicide ideation and attempt (e.g., primary study aims) as a preliminary step in interrogating the assumptions of the ideation-to-action framework. The most consistent group differences across samples were between White/European American and Native American youth. We found that White/European American youth were at consistently lower odds of reporting suicide ideation and attempt across all analyses (e.g., when co-occurring and when occurring separately) compared to Native American youth. This is a particularly important area of contrast, given that the ideation-to-action framework is based on research in primarily White/European American samples. These results may suggest that Native American youth are particularly likely to move along the dimension of suicide severity from thoughts to behavior.

Particularly salient to the question of relevance of current suicide risk models is the preliminary evidence against the ideation-to-action framework. That is, Native American and Black/African American youth are more likely to report the occurrence of suicidal behaviors outside of the context of suicide ideation. To better understand this phenomenon for Native American youth and their minoritized peers, it is critical to consider factors impacting reporting patterns. For example, stigma likely influences suicide disclosure among different cultural groups. However, stigma or suicide reporting hesitancy fails to explain the patterns of reporting suicide attempt without reporting suicide ideation, as suicidal behaviors may be inherently more stigmatized. An alternative explanation is that these reports highlight a form of response bias. In brief, the use of a single-item screening measures for suicide ideation and attempt may lack robustness. That is, if suicide ideation is not conceptualized by youth as “seriously considering attempting suicide” or “thoughts of killing yourself” as indicated in the single-item screener question, youth who have experienced some form of ideation may fail to endorse this question (Ammerman, Burke, Jacobucci, & McClure, 2021). However, when taking a step back and considering the findings in total, there are no clear theoretical reasons to suggest that the differences between Native and White/European American youth in this pattern of reporting would be explained by differences in interpretation of the survey items. Thus, we arrive at the hypothesis that some Native American youth, and in similar proportions other racial/ethnic minoritized youth, may not follow the ideation-to-action framework. As such, it becomes critical to evaluate the clinical implications of these findings.

The assumptions of the ideation-to-action framework are relevant to risk assessment and intervention. Clinical practice commonly relies on the ability to monitor suicidal thoughts and behaviors and then tailor intervention toward the individual’s assessed risk for suicide (e.g., Wiglesworth, 2022). While risk is typically considered along the ideation-to-action continuum, many adolescents who die by suicide have no known history of suicidal ideation or self-harm (Rodway, Tham, Turnbull, Kapur, & Appleby, 2020). Further, we found that those who reported suicide attempt without suicide ideation were approximately 62-81% less likely to report depressive symptoms, a pattern which appears to be mixed in past literature on impulsive suicide attempts (Simon et al., 2002; Weyrauch, Roy-Byrne, Katon, & Wilson, 2001). Further, risk assessment in a clinical setting may be particularly vulnerable to reporting biases above and beyond those demonstrated through anonymous survey data, as youth may be driven to avoid unwelcome follow-up procedures (Fox et al., 2021). In other words, data in this study may undersell the proportion of youth that do not disclose suicide ideation prior to making a suicide attempt in a clinical setting. Yet, even in the context of suicide ideation, it is incredibly difficult to determine if and when an adolescent will transition toward suicidal behavior. Considering that the majority of those who die by suicide are thought to have died from their first attempt (Jordan & McNiel, 2020; McKean et al., 2018), it is profoundly important that safety measures and crisis tools are in place for all at risk youth, regardless of reported suicide ideation or intent, to prepare for the possibility that stressors occur and trigger an impulse to engage in suicidal behavior (as suggested in Fox et al., 2021). This issue may also be compounded among Native American youth, given the complicated relations between Indigenous communities and medical science that has led to broad mistrust. Taken together, clinician decision making, particularly with Native American youth, must not rely primarily on these assumptions of a progression of risk.

Contextualizing Reports of Native and Other Ethnoracial Minoritized Youth

When examining differences between Native American youth and their other ethnoracially minoritized peers, the results here were more nuanced. No significant differences were found between Native American and Native Hawaiian/Pacific Islander youth in any patterns of suicide reporting across both samples. In the nationally representative YRBSS, Asian/Asian American, Black/ African American, Hispanic/Latinx, Multiracial (attempt only) youth were 36-56% less likely to report suicide ideation and attempt as compared to Native American youth but did not differ significantly in endorsing a suicide attempt when endorsing suicide ideation (with the exception of Asian/Asian American youth) or endorsing a suicide attempt without endorsing ideation. Conversely, while Asian/Asian American, Hispanic/Latinx, and Multiracial (attempt only) youth in Minnesota were 24-62% less likely to endorse suicide ideation and attempt, as well as 20-55% less likely to endorse making a suicide attempt when endorsing suicide ideation, than Native American youth, there were not significant differences between these groups in reporting a suicide attempt without also reporting ideation. Finally, Black/African American youth in Minnesota, while being 39% less likely than Native American youth to endorse a suicide attempt when endorsing ideation, were 97% more likely than Native American youth for reporting a suicide attempt without ideation.

These differences between samples raise questions about geographic region as critical context. When considering sampling weights, the YRBSS study included a greater proportion of Back/African American and Hispanic/Latinx youth, and a lower proportion of Native American, Asian/Asian American, and White/European American youth compared to the MSS. While there are hundreds of distinct cultural groups within the Native American population, Native American youth in Minnesota most commonly belong to two cultural groups (Dakota/Lakota and Anishinaabe/Ojibwe). Further, as compared to other areas of the U.S., Black/African American youth in Minnesota represent a smaller proportion of the population, are more likely to live in urban regions, and are more likely to be immigrants (US Census Bureau, 2019). Importantly, these regional differences also likely reflect differences in cultures, systems/structures, and the experiences of Native American and Black/African American youth. Differences in suicide risk and suicide mortality have been documented for Native American and Black/African American populations based on region, cultural background, acculturation, racial socialization, and more (discussed in more detail in Wiglesworth, 2022), thus may underlie the differences observed across the two samples.

Understanding heterogeneity within Native American and Black/African American populations appears to be a critical step in producing more nuanced pictures of suicide risk and resilience and, importantly, heterogeneity is not limited to region and culture. Intersectionality highlights that youths’ racial/ethnic identities do not exist in a vacuum, and instead interact with various additional intersectional such as gender, sexual orientation, social class, religion, and ability status to shape their experiences (and thus their experiences of suicide risk and resilience; Wiglesworth, Clement, Wingate, & Klimes-Dougan, 2022). Future mixed-methods and qualitative research may be particularly well suited for examining how such heterogeneity within different ethnoracial groups, and in particular Native American youth, shapes lived experiences that ultimately alter suicide risk and resilience pathways.

Mechanisms Underlying Engagement in Suicidal Behaviors

While the focus of this study was on identifying patterns of suicidal thoughts and behaviors, a fruitful avenue of future research would be to use a developmental stress lens to delineate mechanisms that might undergird the increased engagement in suicidal behaviors observed among Native American youth. Minority stressors are chronic forms of stress experienced across multiple socioecological systems (e.g., interpersonally, systemically) in overt and covert manners (e.g., overt racism, microaggressions; Cyrus, 2017; Meyer, 2003). While it is likely that youth from all minoritized groups in this study experience minority stressors, the effects of these stressors may be compounded by the experience of other socioecological stressors (e.g., adverse experiences, historical trauma) that Native American youth are documented to experience more frequently than their peers from other ethnoracial backgrounds. It is also important to note the similar suicide reports between Native American and Native Hawaiian/Pacific Islander youth, given the parallels in histories of colonization produce similar ongoing socioecological stressors. Critically, an elevated stress-burden is theoretically linked to the development of biological and cognitive processes that may act as mechanisms that facilitate engagement in suicidal behaviors (e.g., Berger & Sarnyai, 2015), including acquired capability for suicide and impulsivity.

Acquired capability is a key feature of the Interpersonal Theory of Suicide. While thwarted belongingness and perceived burdensomeness are theorized to produce suicidal desire, the acquired capability to harm oneself is theorized as being necessary for one to engage in suicidal behaviors (Joiner, 2005). Critically, painful and provocative events, in addition to dispositional factors and means for self-harm (May & Victor, 2018), increase acquired capability for suicide (e.g., Smith & Cukrowicz, 2010). Relevant to our findings, minority stressors such as discrimination act as painful and provocative events to increase acquired capability, as demonstrated among Black/African American adults (Brooks et al., 2020). Further, adolescents with suicide ideation are more likely to also report engaging in a suicide attempt when demonstrating high levels of acquired capability (Vélez-Grau, Romanelli, & Lindsey, 2022). This same study also found that youth with higher acquired capability are more likely to attempt suicide with no history of suicide ideation. As such, acquired capability, possibly engendered from a history of painful and provocative (e.g., stressful) life events, is likely a mechanism that in part underlies observed patterns of greater engagement in suicide attempts among Native American youth broadly and, moreover, the contributions of minority stressors may in part explain why Native American youth in our study were not more likely than their minoritized peers to report suicide attempts without also reporting suicide ideation. In sum, among those experiencing suicide ideation, acquired capability increases the likelihood of acting in a suicidal manner, and among those without suicide ideation, acquired capability may facilitate life-threatening behaviors that occur in response to an emotional event or impulse (e.g., Miller & Prinstein, 2019), highlighting impulsivity as the second mechanism of interest for future research.

Stressful inputs from the environment are robustly associated with deficits in cognitive (e.g., delayed gratification) and behavioral (e.g., response inhibition) impulsivity. Experiences of harshness (e.g., abuse, overt racism) and unpredictability (e.g., economic insecurity, unsafe environments) are associated with discounted valuations of future rewards, in favor of immediate gratification across development (Frankenhuis, Panchanathan, & Nettle, 2016). In other words, for those whose future is perceived to be in jeopardy, investment in the future is not warranted (e.g., Kidd et al., 2013). Further, these environmental inputs are thought to impact the developing brain, particularly in circuits related to executive function, emotional processing, and impulse control (Berger & Sarnyai, 2015; McLaughlin, Weissman, & Bitrán, 2019), including the prefrontal cortex (Jones & Graff-Radford, 2021) and subcortical structures (Jankowski et al., 2017). These alterations in cognition and biological development following significant exposure to stress are associated with altered stress responses, behavioral impulsivity, and risky behaviors (R. T. Liu, 2019; S. Liu, Cui, Duprey, Kogan, & Oshri, 2020; Oshri, Liu, Duprey, & MacKillop, 2018; Simpson et al., 2012). Thus, impulsivity may promote maladaptive responses to stress among Native American and other oppressed youth, which take the form of suicidal behaviors that prioritize the immediate removal of negative feelings over long-term wellbeing. In a study of Apache adolescents who had previously attempted suicide, 92% reported impulsive suicide attempts (Cwik et al., 2015). In other populations, such impulsive attempts have been recorded to occur less often, at an estimated 24–74.4% of those who attempt suicide (Deisenhammer et al., 2009; Paashaus et al., 2021; Simon et al., 2002). Thus, this impulsivity hypothesis provides a potential underlying mechanism that in part explains the reporting of suicide attempt in absence of, or above and beyond, suicide ideation as also noted by others (Bolton et al., 2014; LeMaster et al., 2004). Moreover, male youth are overrepresented among impulsive suicide attempters (Simon et al., 2002; Weyrauch et al., 2001) and similarly were overrepresented among those who reported suicide attempt without suicide ideation in our study and in previous research (Romanelli, Sheftall, Irsheid, Lindsey, & Grogan, 2022).

Limitations

There are a number of strengths in this study including, most notably, the use of two large, epidemiologically informed assessments of adolescents across the U.S. (YRBSS) and the state of Minnesota (MSS). Additionally, generally consistent approaches were employed to measuring key outcomes of suicidal ideation and attempts. Nevertheless, this project is merely a starting place for raising the possibility that the ideation-to-action framework is not equally relevant for youth across all ethnoracial groups. This work has a number of limitations, some of which are common across studies that make use of large, health-risk screening surveys such as those used here. First, longitudinal research is needed to fully address the questions of how suicide risk unfolds over time. The use of validated, reliable scales for examining suicidality (e.g., Keane et al., 1996) administered prospectively along with the inclusion of risk and protective mechanisms would lend to a more robust understanding of suicide risk progression during adolescence. Retrospective reports are also subject to recall bias, which we hoped to in part limit by examining past year reports of suicide ideation/attempt as opposed to lifetime (Klimes-Dougan, Mirza, Babkin, & Lanning, 2022). In considering future directions, this line of inquiry would also benefit from more granular assessments of suicidality (e.g., including passive ideation, planning, preparatory behaviors). Second, while the samples are large (particularly the MSS), some of the ethnoracial groups are small in size considering the rarity of our outcomes of interest (e.g., Native Hawaiian/Pacific Islander youth, Native American youth in YRBSS), making it difficult to determine if the lack of replication across samples is due to limited power. Third, the coding of ethnoracial identity is important to consider when interrogating external validity. While the CDC and U.S. Census routinely separates multiracial individuals into a distinct category, this approach has limited utility given the vast heterogeneity of cultural identities within and the growing multiracial population in the U.S. (Cho, Crenshaw, & McCall, 2013; Trimble, 1995). Fourth, there are a number of measurement considerations, including the use of dichotomous, retrospectively reported screening variables that limits nuanced interpretations of these findings and prevents assessing for construct and internal validity. Moreover, recoding of the depression variable was necessary to match across years in the MSS, which is not ideal but was necessary for increasing our sample size. Fifth, both surveys provided limited demographic variables that were able to be included as covariates in the analyses. We were unable to account for geographic region (e.g., rural, urban, suburban, reservation), Tribal affiliation, and socioeconomic status, as just a few examples. Sixth, careful consideration is needed when including cohorts of participants surveyed at different timepoints. For this archival data it is important to note that suicide risk reporting was surveyed prior to the onset of the Covid-19 pandemic, which has disproportionately affected Black/African American, Native American, and Hispanic/Latinx communities in the U.S. (Yellow Horse & Huyser, 2022) and had broad impacts on adolescent mental health (Bridge et al., 2023; Panchal et al., 2021). Finally, while our a priori hypotheses were reported, they were not pre-registered.

Conclusion

The current project, when viewed in tandem with existing literature, posits that Native American adolescents are both more likely than youth from other ethnoracial backgrounds to think about and engage in suicidal behaviors and more likely than their White/European American peers to endorse suicide attempt with and without endorsing suicide ideation. While these data suggest that Native American youth are more likely than their peers to progress from ideation to action, they also provide preliminary evidence that refutes the fundamental assumption of the ideation-to-action framework that suggests ideation is required prior to attempt. This work ultimately emphasizes the importance of understanding mechanisms that facilitate engagement in suicidal behaviors among Native American youth. Understanding the potential patterns of suicide risk trajectories is a critical need in future research and should inform clinical intervention efforts that depend on our ability to accurately gauge risk for engagement in suicidal behaviors and intervene effectively.

Abbreviations:

YRBSS

Youth Risk Behavior Surveillance Survey

MSS

Minnesota Student Survey