Trajectories of Well-being among Latina Adolescents who Attempt Suicide: A Longitudinal Qualitative Analysis

Abstract

Significant research questions persist regarding the short- and long-term outcomes of Latina adolescents who attempt suicide. To address these limitations, we utilize an ecodevelopmental framework to identify potential factors that shape differential outcomes following a suicide attempt. Through an exploratory, longitudinal, qualitative research design, we investigate two research questions: How do trajectories of well-being vary among Latina teens after a suicide attempt? What risk and protective factors might contribute to different trajectories? We conducted qualitative interviews with 17 Latina participants living in predominantly low-income households in New York City. Interviews took place within the six-months following their suicide attempts, and again, 12 months later. Analysis revealed three distinct trajectories after a suicide attempt: resilience, tenuous growth, and chronic stress. Our findings elucidate potential factors that contribute to resilience following a suicide attempt and underscore the importance of prevention and intervention programs that foster adolescents’ connectivity across ecodevelopmental contexts.

Research has consistently shown that Latina girls living in the U.S. experience high risk for suicidal behaviors compared to non-Hispanic White, Black, and Asian American adolescent females (Romero, Edwards, Bauman, & Ritter, 2014; Price & Khubchandani, 2017). In 2017, 8.2% of Latina teens surveyed (N=2,796) reported a suicide attempt (Centers for Disease Control and Prevention [CDC], 2018). Of significant concern is the fact that Latina teens often engage in multiple suicide attempts, ranging from two to six attempts during adolescence (Hausmann-Stabile, Kuhlberg, Zayas, Nolle, & Cintron, 2012). The high prevalence and recurrence of suicidal behaviors has significant—and sobering—implications for the future well-being of Latina youth. For some adolescents, non-lethal suicidal behaviors may lead to a trajectory involving future suicidal ideation, suicide attempts, and adult psychopathology (Goldman-Mellor et al., 2014). Additionally, non-lethal suicidal behaviors are strong predictors of completed suicide (Wilkinson, Kelvin, Roberts, Dubicka, & Goodyer, 2011).

A small, but growing, body of suicide research has focused on short- and long-term outcomes of adolescents who attempt suicide. Few of these studies have focused exclusively on Latina adolescents. In the U.S., Latina girls represent a large population that encompasses considerable diversity in terms of immigration experiences, legal status, socioeconomic status, acculturation, and national heritage (Romero et al., 2014). Scholars posit that within this diverse population, certain factors—such as migration, acculturative stress, socioeconomic stressors, ethnic discrimination, and family conflict—increase risk for suicidal behaviors (Duarté-Vélez & Bernal, 2007). To date, few data exist to evaluate the potential negative impacts of suicidal behaviors on the future well-being of Latina teens. The purpose of this exploratory, longitudinal, and qualitative study is to examine potential trajectories of well-being following a suicide attempt among Latina teens living in predominantly low-income households in New York City. Establishing pathways for well-being following a suicide attempt is an important first step toward the development of effective responses to reduce rates of repeated suicidal behaviors among Latina teens.

Longitudinal Outcomes of Youth following a Suicide Attempt

Most studies on adolescent outcomes following a suicide attempt have focused on the risk of future suicide attempts. Studies conducted with community samples indicate that adolescents who attempt suicide are three to eight times more likely to reattempt compared to youth with no lifetime history of suicidal behaviors (Wichstrøm, 2000; Lewinsohn, Rohde, Seeley, & Baldwin, 2001). Risk for recurrent suicidal behaviors is greater in clinical populations, and several studies show that between 10% and 20% of adolescent psychiatric inpatients will reattempt within 18 months following hospital discharge (Spirito et al., 1992; Brent et al., 1993; Goldston et al., 1999; King et al., 1995; Prinstein et al., 2008). Higher repetition rates appear in studies with longer follow-up periods: up to 50% of adolescents reattempt within the three years following their index suicide attempt (Boergers & Spirito, 2003).

Researchers recognize repeat suicide attempts to be a salient risk factor for eventual death by suicide (Christiansen, & Frank Jensen, 2007; Bostwick, Pabbati, Geske, & McKean, 2016; Hawton & Fang, 1988), and several scholars have directed attention to identifying those factors that increase the likelihood for future suicidal behaviors. Adolescents who reattempt suicide are more likely to report problems in psychosocial functioning, including a high prevalence of serious mental health problems (Fergusson, Horwood, Ridder, & Beautrais, 2005; Groholt & Ekeberg, 2009), and continued suicidal ideation (Goldston et al., 2016; Spirito, Valeri, Boergers, & Donaldson, 2003; Yen, Weinstock, Andover, Sheets, Selby, & Spirito, 2013; Wolff et al., 2017). For example, Czyz and King (2015) examined trajectories of suicidal ideation among psychiatrically hospitalized adolescents who attempted suicide. Sub-clinical suicidal ideation distinguished the two most common trajectories in the sample: in both groups, there was a rapid decline in suicidal ideation within three months following hospital discharge. The least common trajectory group exhibited high suicidal ideation at hospitalization, followed by persistent and elevated suicidal ideation over the course of the year after hospital discharge. Risk for subsequent suicide attempts was significantly more likely to occur among adolescents with severe and chronic suicidal ideation (Czyz & King, 2015).

Despite a burgeoning literature on adolescent outcomes following a suicide attempt, several gaps in knowledge remain. Research to delineate psychological risk factors has advanced much farther than efforts to understand the interpersonal and social factors that shape continued suicide risk following a suicide attempt (Goldman-Mellow et al., 2014; Mars et al., 2014). Additionally, very little research focuses on potential protective factors. In one noteworthy study, Czyz and colleagues found that improved family dynamics following a suicide attempt decreased the likelihood of severe suicidal ideation, depressive symptoms, and future suicidal behaviors (Czyz, Liu, & King, 2012). The lack of research focusing exclusively on the experiences of Latina teens leaves unresolved the important questions of how trajectories of well-being vary among Latina teens who have attempted suicide, and which factors potentially shape those trajectories on their pathways toward recovery. To address these questions, we utilize an ecodevelopmental framework to identify and analyze how various experiences, contexts, and situations potentially shape differential outcomes following a suicide attempt.

An Ecodevelopmental Approach

An ecodevelopmental approach provides a comprehensive framework that evaluates individual behaviors within an interactional web of nested systems: micro-, meso-, exo-, macro-, and chronosystem (Bronfenbrenner, 1979; Tudge, 2008; Ungar et al., 2013). Micro-level analysis addresses those relationships with which an individual has direct contact, such as family members, peers, schools, and people in the community (Coatsworth et al., 2002). The meso-level describes links between microsystems, such as how an adolescent’s interactions with her mental health therapist shape interactions with her parents (Ayón, Marsiglia, Bermudez-Parsai, 2010). The exo-level includes those processes that contribute to the organization of relationships at the meso- and micro-level, emphasizing that an adolescent’s well-being can be affected by settings with which she has no contact. Examples can include a parent’s experiences with the criminal justice system or immigration enforcement (Ayón, 2016). The macro-level refers to broader institutional systems (e.g., cultural, economic, or political systems) and ideologies (e.g., racism, xenophobia), and the ways in which macrosystems manifest through linkages with family, peers, and school or work settings to shape well-being (Ayón, Ojeda, & Ruano, 2018). The final component, the chronosystem, emphasizes the importance of change at each level over time and how changes and transitions shape individual outcomes (Baldwin-White, Kiehne, Umana-Taylor, & Marsiglia, 2017).

To date, much of the research that applies an ecodevelopmental approach to the study of suicidal behaviors among Latina teens has prioritized interactions at the micro-level. Such work references an important component of Bronfenbrenner’s ecodevelopmental theory: proximal processes. Bronfenbrenner argued that proximal processes, or “enduring forms of interaction in the immediate environment,” are the most salient mechanisms shaping youth developmental outcomes (Bronfenbrenner & Morris, 1998, p. 996). In research on adolescent suicide risk, studies have emphasized the role of family conflict, family violence, stressed family relationships, and family dysfunction in shaping the onset of suicide attempts among Latina adolescents (Razin et al., 1991; Garcia, Skay, Sieving, Naughton, & Bearinger, 2008; Gulbas & Zayas, 2015; Herrera, Dahlblom, Dalgren, & Kullgren, 2006; Zayas, Gulbas, Fedoravicius, & Cabassa, 2010). Additional research has demonstrated how contrasting intergenerational cultural values about parenting style, adolescent autonomy, and the allocation of household rules and responsibilities contribute to family conflict (Cardona, Nicholson, & Fox, 2000; Szapocznik & Williams, 2000; Halgunseth, Ispa, & Rudy, 2006), and can precipitate a suicide attempt (Zayas et al., 2010).

Despite the salience of the micro-level in shaping suicide risk, it is important to consider how experiences and opportunities at the micro-level develop within broader contexts, as Bronfenbrenner originally intended. Failing to acknowledge ways in which broader processes “trickle down” to shape micro-level experiences hazards a misrecognition of the impact of larger structural process on mental health and well-being over time (Tudge, Mokrova, Hatfield, & Karnik, 2009). Considering this, we apply an ecodevelopmental framework, in its complete form, to identify and describe processes that both constrain and provide opportunities for well-being in the aftermath of a suicide attempt. To foreground Latina adolescents’ understandings and descriptions of their experiences, we utilize an exploratory, longitudinal, qualitative research design to investigate two research questions: How do trajectories of well-being vary among Latina teens after a suicide attempt? What risk and protective factors might contribute to different trajectories?

Methods

Recruitment

The research team recruited participants from 2005 to 2010 as part of their involvement in a large federally-funded, mixed-method project to examine the influences of sociocultural process on the suicidal behaviors of Latina teens living in predominantly low-income households in New York City. A total of 73 Latinas who attempted suicide participated in the qualitative phase of the large study. From this sample, 17 participants agreed to take part in a follow-up interview, which took place approximately one year following the baseline interviews. Primary reasons for attrition from the initial participant population were loss of contact through disconnected phone numbers and housing instability.

The research team recruited participants from multiple sites, including a municipal hospital that provided psychiatric emergency and outpatient services, a private psychiatric hospital, and communities agencies that provided outpatient mental health services. Research personnel worked closely with agency staff and mental health clinicians, who identified potential participants and obtained their permission to be contacted by the research team. Agency staff also assessed each girl’s readiness to participate to ensure that each participant had the capacity to choose freely to participate without the presence of coercion or potential mental health risks.

To participate in the study, adolescents needed to be residing with their families at the time of a suicide attempt occurring no more than six months prior to the baseline interview. Exclusionary criteria for participants included diagnosis of a major mental illness (e.g., schizophrenia, bipolar disorder) or intellectual disability because these conditions present unique challenges thought to shape the etiology and prognosis of suicidal behaviors in unique ways (Hauser, Galling, & Correll, 2013; Fujita et al., 2015; Martin, Thomas, Andrews, Hasking, & Scott, 2015; Mayes, Calhoun, Baweja, & Mahr, 2015). All institutions where research activities took place provided IRB approval, and all participants and their caregivers gave assent and consent for their participation. During the informed consent process, interviewers informed participants that they would be compensated for their participation in the interview. Participants received compensation prior to the interview to avoid the potential for financial coercion (e.g., feeling compelled to answer a question to receive compensation; Kaiser, 2009).

Participants

In this article, we include only those participants who completed both baseline and follow-up qualitative interviews for analysis (n = 17; see Table 1). The ages of participants ranged between 14 and 18 (inclusive), with an average age of 16 years. The majority of youth were born in the United States (76.5%), and represented several different national heritages, including Puerto Rican (n = 5), Dominican (n = 6), and Mexican (n = 4).

Table 1.

Characteristics of Adolescent Latinas in Sub-Sample

	Total participants (N=17)
	M ± SD	n (%)
Age in years
11-13	13.1±. 45
14-16	15.3 ± .89
17-19	17.47 ±. 60
US Born		13 (76.5)
Legal Status
Citizen		14 (82.4)
Resident		2 (11.8)
Undocumented		1 (5.9)
Hispanic Cultural Group
Puerto Rican		5 (29.4)
Dominican		6 (46.2)
Mexican		4 (23.5)
Other1		2 (11.8)

Notes:

¹This includes participants who identified as Ecuadoran or Venezuelan.

Qualitative Interviews

Interviewers conducted the qualitative interview in either English (n = 14) or Spanish (n = 3), based on the participant’s preference. The team implemented several measures to facilitate rapport and promote participant safety over the course of the project (Jordan et al., 2012). The team comprised bicultural, bilingual, masters-level clinicians with degrees in social work or related fields. Clinical interviewers did not provide clinical or therapeutic services to participants before or during the project, although they did provide a list of psychological and behavioral health services and facilitate referrals to address unmet needs for mental health services. All interviewers received training to conduct in-depth qualitative interviews to ensure that: 1) interviewers posed questions and follow-up probes in ways that elicited detailed responses; and 2) participants felt comfortable to talk freely and openly about their experiences.

To facilitate consistency in the topics explored across interviews, interviewers used a semi-structured interview guide. Baseline questions explored the participant’s perception of family dynamics and relationships; social connection and experiences outside of the family and home (e.g., school, neighborhood); and a detailed, retrospective account of the suicide attempt. Follow-up interviews, which occurred one year following baseline interviews, focused primarily on how the participant’s life experiences had changed since the baseline interview, and included questions about mental health and well-being, family, peers, school, neighborhood and community, and other domains pertinent to the participant’s life. Project personnel transcribed recordings of each interview in the original language. Interviews averaged 50 minutes in length.

Data Analysis

Given our interest in understanding and analyzing patterns in trajectories of well-being from an ecodevelopmental perspective, we utilized a thematic approach that combined deductive (theory-driven) and inductive (data-driven) analysis (Fereday & Muir-Cochrane, 2006; Guest, MacQueen, & Namey, 2013). Thematic analysis represents a flexible method aimed at identifying, organizing, and analyzing themes across a qualitative data set (Simon, Lathlean, & Squire, 2008). Unlike other common approaches in qualitative research, including grounded theory, discourse analysis, and narrative analysis, thematic analysis is not associated with a specific theoretical tradition and can be applied across a range of epistemological orientations, including inductive and deductive analysis (Nowell, Norris, White, & Moules, 2017).

In line with our theoretical approach, we developed a list of codes that focused on risk and protective factors situated within various levels in the ecodevelopmental model, including micro- (e.g. family, peers), meso- (e.g. school, community), and exo- and macro-level (e.g. criminal justice system, discrimination). To develop new codes and evaluate the codebook within participants’ narratives, the research team read baseline and follow-up interviews associated with five participants. During a team meeting, we revised the codebook to include codes such as positive ethnic identity and experiences of violence (see Table 2). Additional transcript readings did not yield any new codes or revisions to codebook definitions, suggesting that we had reached code saturation (Guest et al., 2013).

Table 2.

Qualitative Themes that Emerged in Interviews with Latina Teens who Attempted Suicide (n = 17)

Code	Definition	Example
Protective Factors
Family Connection	References events, feelings, activities, and circumstances that indicate and/or promote warmth, intimacy, support, and togetherness within the family, including ways participants feels supported by individual family members.	“They do support me. They tell me to keep going, to achieve … not to dream, just achieve what I want.”
Service Provider	A statement, event, or feeling that demonstrates support with a therapist or case manager.	“Like the only one that I really felt like close to, and the only one that I could like share my feeling with, was my therapist at the time.”
School Environment	An explicit reference to the ways the participant feels supported, connected, and/or engaged in the school community, including relationships or interactions with school staff such as teacher, police officer, principal	“This school’s better for me because they won’t have the problems I don’t want and need.”
Positive Ethnic Identity	Descriptions of pride, strength, and connection to participant’s ethnic identity	“When you represent your flag. That’s when I feel most Latina. We have pride. We speak out. We don’t let nobody bring us down or nothing.”
Risk Factors
Socioeconomic Strain	A description of parent’s (un)employment status, inability to meet subsistence needs, housing insecurity, or ways in which family members find ways to supplement household income.	“I was working there when I was 12 ‘cause my mom at the time, she was needing help, like with the bills and everything. So, I worked in Jersey.”
Neighborhood Risks	References feelings of fear, lack of safety, worry, stress, sadness, etc., that stem from neighborhood environment, in addition to experiences of risks at the neighborhood level, such as burglary, shootings, violence, or gang activity.	“Like we wouldn’t go outside as much. We wouldn’t go outside, or if one of us did go outside we would go as a group. We would never leave the house alone, stuff like that.”
Criminal Justice System	Can refer to a participant’s or family member’s involvement in the criminal justice system, including circumstances, feelings, perception, and experiences.	“The cops came to arrest him ‘cause they found out he had drugs in the basement. He got arrested, and he might get deported. And ever since then I haven’t seen him.”
Violence	Description of experiences of physical or sexual abuse, or witnessing domestic violence	“My stepfather did hit me, and I went to school the next day with a bruise on my face.”

Notes. This table does not reflect all themes identified in analysis of interviews with participants and described in the project codebook. Additional themes associated with protective factors included perceptions of self-worth, coping skills, peer support, and community support. Additional themes associated with risk factors included family risk factors, barriers to mental health services, discrimination, peer victimization, cognitive vulnerabilities, risky behaviors, and school challenges.

Using NVivo10, two team members double-coded all transcripts. We discussed incongruences or questions in the application of codes during team meetings, which focused on evaluating the coded text in question, analyzing the text in relationship to the whole transcript, and reaching consensus on the code application (Bradley, Curry, & Devers, 2007). After coding, we entered coded data into a time-ordered matrix (Miles, Huberman, & Saldana, 2014), organizing columns in the matrix by code at baseline and follow-up, and rows by participant. Cells contained reduced and summarized information, code examples, and illustrative participant quotes to describe how each code, at a particular time point, manifested in the participant’s interview.

The research team analyzed the matrix to identify thematic patterns across participants in the sample. Specifically, we used the matrix to identify adolescents who shared similar experiences of change across time (Kaimal & Beardslee, 2010), grouping participants who shared thematic configurations together as illustrating a specific “trajectory.” From this, we wrote a “composite summary” (Shank, 2006) for each participant to describe and situate trajectories based on the coded data. A research team member independently evaluated composite summaries to validate each participant’s designation within a particular trajectory.

To develop the findings presented below, we applied an iterative process of working across the time-ordered matrix, participant composite summaries, and individual interview transcripts. We analyzed similarities and differences within and across trajectories to gain insight and contextualize our interpretations about factors that held potential to shape trajectories. To limit the potential for bias and integrate alternative interpretations of the data, we discussed our explanations of each case during a team meeting (Ryan & Bernard, 2000).

Results

Analysis of the qualitative interviews revealed three distinct trajectories in the aftermath of a suicide attempt: resilience, tenuous growth, and chronic stress. We defined these trajectories based on how participants explained their acts of self-harm and stressors at the baseline interview; and how risk and/or protective factors changed or remained the same 12 months post-suicide attempt. The resilience trajectory emphasizes the availability of supports and resources across psychological, interpersonal, and sociocultural domains that promoted and sustained well-being following a suicide attempt. Tenuous growth reflected a general trajectory toward psychosocial well-being, but in a context of the sustained presence of multiple risk factors that continued to affect recovery. The third trajectory, chronic stress, described the continuation of stressors and vulnerabilities across domains in the adolescent’s life, including persistent suicidal ideation and/or re-hospitalization.

Below, we explore each of these trajectories more broadly to illustrate the complex interplay between risk and protective factors and how this interplay resulted in different trajectories of well-being following a suicide attempt. Following this discussion, we draw on several instrumental case examples to describe each trajectory in more detail. We do not intend the presentation of these cases to generalize across cases, but rather to lend qualitative richness to our general understanding of processes of change or stagnation that might shape recovery after a suicide attempt (Gaikwad, 2017).

Trajectories of Well-being

At baseline, all participants, regardless of their trajectory, reported several risk factors that they perceived as being relevant to their suicide attempt (see Table 3). Risk factors, such as mother-daughter conflict, family separations, histories of self-harmful behaviors, experiences of domestic or physical violence, mental health problems, and socioeconomic stress, manifested among participants across the three groups. However, participants coded in the tenuous recovery and chronic stress trajectories were more similar in terms of the configuration of risk factors at baseline. As described in Table 3, girls in these two trajectories had additional risk factors, including father-daughter conflict, experiences of sexual violence, engagement in risky behaviors (e.g., substance use, cutting school), and housing insecurity. Girls in the tenuous recovery and chronic stress trajectories did not report any protective factors, whereas girls in the resilient trajectory referenced the availability of social support at baseline.

Table 3.

Factors that Shaped Trajectories of Well-being over Time

Trajectory	Baseline Interview	Changes at Follow-Up Interview	Psychological Well-Being
Resilience	Conflict with mother Cognitive vulnerabilities (e.g., hopelessness, perceived worthlessness) Family separations History of self-harmful behaviors Mental health problems Socio-economic strain Supportive community (e.g., school, church) Positive relationship with father	Improved relationship with mom Development of a positive ethnic identity Development of positive and effective coping skills Moving to a new, positive school environment	Reduction in cognitive vulnerabilities (e.g., hopelessness, perceived worthlessness) Emergence of cognitive strengths (e.g., positive self-esteem, future aspirations) Absence of suicidal ideation Absence of engagement in self-harmful behaviors
Tenuous Growth	Conflict with mother Cognitive vulnerabilities (e.g., hopelessness, perceived worthlessness) Family separations History of self-harmful behaviors Violence in the home Mental health problems Socio-economic strain Conflict with father Housing insecurity Sexual violence Cutting school Neighborhood violence	Engagement in new risky behaviors (e.g., risky sexual activity, cutting school, fighting in school) Development of a positive ethnic identity Development of positive and effective coping skills Moving to a new, positive school environment	Reduction in cognitive vulnerabilities (e.g., hopelessness, perceived worthlessness) Emergence of cognitive strengths (e.g., positive self-esteem, future aspirations) Absence of suicidal ideation Absence of engagement in self-harmful behaviors
Chronic Stress	Conflict with mother Cognitive vulnerabilities (e.g., hopelessness, perceived worthlessness) Family separations History of self-harmful behaviors Violence in the home Mental health problems Socio-economic strain Conflict with father Housing insecurity Sexual violence Cutting school Neighborhood violence	Engagement in new risky behaviors (e.g., risky sexual activity, cutting school, fighting in school)	Cognitive vulnerabilities (e.g., hopelessness, perceived worthlessness) Suicidal ideation Re-hospitalization

Notes. Protective factors are identified in italics.

At follow-up, all participants expressed considerable changes in their everyday lives following their suicide attempts. Participants in the resilience and tenuous growth trajectories discussed the emergence of protective factors, including positive ethnic identity, effective coping strategies, and changes in their school environment; which helped them move toward a state of psychological well-being. In addition to protective factors, girls in the tenuous recovery group experienced the emergence of new risk factors—that is, the presence of a new risk factor that was not reported at baseline—such as experiences with the criminal justice system. Girls categorized in the chronic stress trajectory also cited the emergence of additional risk factors. However, unlike participants in the resilience and tenuous recovery trajectories, they expressed a continuation of suicidal ideation and hopelessness at follow-up, and they often attributed this to the lack of protective factors in their lives.

Resilience Trajectory

In our sample, we categorized five participants in the resilience trajectory as they exhibited protective factors at both baseline and follow-up. Although they manifested fewer protective factors at baseline, these participants reported improvements in their own self-perceptions, coping skills, or relationships with family members and peers over time. Often, improvements in one domain, such as the development of new coping skills, helped to facilitate improvements in other domains, such as relationships with family. Links across domains suggest an interactive and dynamic process of resilience following a suicide attempt, which is evident in the following case.

In this case, the participant’s trajectory toward resilience centered on becoming connected, both with herself and others in her social network. Prior to the attempt, the participant moved to the United States from Latin America to live with her mother, who had immigrated many years prior. The participant described the transition as particularly difficult: “So, like, when I came here, I was very depressed ‘cause I didn’t see the people that I used to see before. Like I lived with different people that I hadn’t since I was like months old.”

Although the participant “learned to get along” with her new family, she continued to experience emotional vulnerabilities, including intense and overwhelming feelings of worthlessness and hopelessness. As she explained, “I felt like nothing. I felt as if every time I tried, I wouldn’t succeed. So why do it?” The participant rarely talked with her family or friends about how she was feeling. She was particularly wary of expressing her emotions in front of her mother, whom she described as “the type of person you can’t tell your emotions to. You can’t say how you feel. You can’t express how you feel. And if she asks, you feel scared to tell her because you’re scare she would not like it.”

On the day of her suicide attempt, the participant wrote a note, a kind of “poem telling myself how I tried committing suicide.” After a teacher discovered the note, the participant was referred to treatment. Initially, the participant struggled to manage her intensive therapy sessions, both inpatient and outpatient. Yet, over time, she realized her work in therapy was helpful. She reflected, “It was weird, because I never thought I would get, like, cared for the way I felt. I didn’t know if it was gonna help me or not. But, it’s been fun.” Over time, the participant reported improvements in her feelings of low self-esteem and depression, in addition to her connections with people in her social network. As she summarized her experience,

“Things are going better. My depression is not that low anymore, and I have more esteem to myself. I got a lot of the people around me, part of why I’m not so depressed. I used to think like people didn’t think I was there, or I felt like I wasn’t important. But now, it’s a little bit different.”

Ultimately, the shift from risk to resilience was an important process shared among all participants classified in the resilience trajectory. Prior to the suicide attempt, each participant pointed to family problems as a central factor to her decision to attempt suicide. Poor forms of communication (e.g., arguing or yelling), disruptions in the family structure through immigration or parental separation, or the lack of a nurturing family environment contributed to feelings of disconnection. In contrast, after the suicide attempt, participants reported a renewed sense of connection to family. For example, one participant noted that “even though there’s problems in the world, there’s still people that care about you. Like, my mom is gonna be really there for me, even though I have a conflict with her.”

In every case, access to mental health services played a crucial role in facilitating connection for participants, suggesting the importance of meso-level factors. Therapy promoted the development of cognitive strengths and coping skills that showed girls, in the words of one participant, how to “get out of their heads and express their emotions.” Emotional expression, which could take the form of journaling or talking to friends and family members, resulted in an appreciation for “just living life the way it is, not always passing by.”

Tenuous Growth Trajectory

In our sample, six participants exhibited tenuous growth. Tenuous growth references both the lack of protective factors at baseline, and the continuation of risk factors—alongside the emergence of protective factors—at follow-up. Importantly, each participant described that feelings of worthlessness and hopelessness had stopped. Yet, in contrast to girls in the resilience trajectory, they continued to experience vulnerabilities across ecodevelopmental contexts, such as continued family conflict, problems with the criminal justice system, neighborhood violence, or discrimination.

In one instrumental case, the participant described her life as really “stressed” prior to her suicide attempt. She explained that multiple stressors at the micro- and macro-level contributed to overwhelming feelings of worthlessness. At the micro-level, she described herself as “really, like, conscious about myself. Like, I always had a low self-esteem. Then my mom and me, being that we didn’t get along. It made me feel like worthless.” Beyond these experiences, she also reflected on the ways in which macro-level factors, including discrimination and marginalization, shaped her perceptions of herself.

“You feel like you don’t have, like, as many chances as white people do. You know, you don’t dress the same way that they do. You don’t talk the way that they do. You don’t do anything that they do. And it’s, like, it makes you feel like crap, you know?”

This participant never spoke with anyone about her feelings, in large part, because she felt that everyone in the family always argued. Due to protracted fights and arguments, the family was in therapy prior to the participant’s attempt. Yet, even then, “I never told anybody about my problems. I’d just sit in like another corner, and listen to everything.” Ultimately, the experiences of multiple vulnerabilities culminated in a suicide attempt. As she remembered,

“I was just mad about everything. Mad at the world, mad at myself. Like, it sucks because after a while, like, you know, you try to be like this perfect person so that, you know, like, make everyone else happy. And you try to make yourself happy, too. But in the end, you feel worthless. Like, I see no point in my life.”

Immediately after attempting suicide, the participant’s relationship with her mother worsened. “[My mom] was like, ‘Oh, I hate you! You are my worst mistake!’ And she was like, ‘I hope you die. I hope I never see you again.’” With support from her therapist and grandmother, the participant made the difficult decision to sever permanently her ties with her mother and live with her grandmother full-time. Still, it was a difficult transition. In reflecting on her relationship with her grandmother, the participant noted, “we don’t agree on much. We don’t agree on anything! But, we try to, you know, make it work.”

Despite continued vulnerabilities within her environment, including family conflict and an escalation in neighborhood gun violence, the participant noted that she no longer felt worthless. “I feel better. Like, my self-esteem isn’t as low as it was before. It’s much better. I’m more open to people. I can share my story. I can count on myself.” The participant credited her personal growth to her individual work in therapy. “I’m like, “How does [the therapist] know this stuff?’ You know? She just wants to be there and help me make my decisions, and sort of guide me through everything. She’s just there, like a helping hand, and someone I can turn to whenever I, like, need the most help.”

The cultivation of hope and development of coping skills were key elements reflected across participants in the tenuous growth trajectory. A growing network of supportive individuals in the lives of Latina youth shaped their ability to have a hopeful perspective. As one adolescent described, “I see how much love people give me. The friends I have, just recently have in school, they show me love every day.” The emergence of protective factors helped to buffer against both new and ongoing stressors in the lives of participants in the tenuous recovery trajectory. In the case of a different participant, for example, a violent encounter in school led to her arrest. With a looming court date, the participant was worried: “I don’t wanna go to jail. They’re going to beat me up in there. I’m very young. I’m like the only young one in there ‘cause they wanna judge me as an adult.” Despite the presence of a salient stressor, the participant maintained an encouraging outlook. As she proclaimed, “Life is happening for a reason, just think about it in a positive way.” Indeed, this trajectory—with its emphasis on hope in the face of daunting challenges—was best summed by another participant, who exclaimed, “There may be some downfalls, [but] there is always a rainbow at the end.”

Chronic Stress Trajectory

We classified six participants in our sample in the chronic stress trajectory. All participants in this trajectory continued to report suicide-related risk factors at follow-up, including depressive symptoms, hopelessness, suicidal ideation, or re-hospitalization. In one case, for example, the participant described her life in terms of “stress” and “pain.” During her middle childhood, the participant experienced three years of chronic sexual abuse by a family member. She finally disclosed the abuse to her mother, who ignored her pleas for help. In reflecting on that moment, the participant noted,

“She didn’t do anything. I expected her to do something; she didn’t. And, I felt like she, maybe she didn’t do it ‘cause she was in, like, denial. I don’t know what it was. I was frustrated that she didn’t do anything, so I saw that she didn’t care. That’s why we don’t get along now.”

Approximately two years after the abuse ended, the participant began to cut herself. Although she experienced temporary emotional relief through cutting, challenges in the broader ecodevelopmental environment continued to emerge in her everyday life. She began to fear that the family would get evicted, explaining, “We’ve had to deal with [that] in the past. So, I’m really scared of that.” She also became pregnant, but miscarried soon after. She felt the pregnancy momentarily improved her relationship with her mother.

“Like, when I was pregnant, my mother was there for me. And everything was fine. But then, after I lost the baby, it’s like she didn’t care anymore. She was like, ‘since there’s no baby, I don’t have to, like, be there for you as much anymore.’”

The feelings of loss and loneliness soon overwhelmed this participant, and cutting no longer provided the relief that it once did. Although she was in therapy, the participant explained that she did not trust her therapist, and she “didn’t want to continue with her.” On the day of her suicide attempt, she explained “nobody is really there. There’s nobody I could talk to. I just keep everything inside.”

Nearly one year after the first interview, the participant noted that “it’s difficult to live” her life. Her relationship with her mother continued to be strained, and the neighborhood where she lived had become more dangerous and violent. “There’s not a way we can really protect ourselves. It’s like, there’s a code to live there. If you do something that doesn’t follow that code, you gotta deal with the consequences.” Despite the continuation of numerous stressors, she expressed that she was committed to stopping her cycle of self-harm. However, it was often difficult. As she explained,

“My therapist, she tries so many different things to, like, to try to get me from not cutting myself. She would tell me, ‘try writing it out, like just writing and expressing yourself. Maybe you could try calling a friend and talking to them, or just having somebody there.’ The friend-thing, it didn’t really work, ‘cause I still cut myself. Now, since I write a lot, and I enjoy writing, it helps more.”

This participant described that outside of her relationship with her therapist, she did not talk with anyone else about her feelings or experiences. When asked to describe the year following her suicide attempt, she summarized by saying “it’s been really stressful and hard to cope with everything.”

Other participants in the chronic stress category reflected similar expressions in their narratives. For example, one adolescent described her continued struggle with her mental health: “I have times when, like, I get like so depressed that I feel like I’m going insane. Like, you know, honestly, I don’t, I don’t really know what is wrong with me. I don’t know, but I feel like I’m dead.” When asked how life had been in the year following the baseline interview, another participant stated, “bad, really bad.” Due to enduring struggles, coupled with few avenues for support, participants in the chronic stress trajectory had a difficult time envisioning a positive future.

Discussion

This article informs a growing body of literature on longitudinal outcomes of suicidal behaviors among Latina teens. Guided by an ecodevelopmental approach, we explored adolescent well-being in the aftermath of a suicide attempt to understand risk and protective factors that shaped trajectories over time. Through the examination of multi-level interactions within and across participants’ lives, we identified three different trajectories post-suicide attempt: resilience, tenuous growth, and chronic stress. The complex interplay of risk and protective factors over time shaped categorization within a specific trajectory. Our findings show that adolescents experienced a range of risk factors from micro-level factors, such as psychological vulnerabilities and family conflict, to broader exo- and macro-level factors, including discrimination, housing insecurity, and neighborhood violence.

The case studies described above highlight the ways in which risk factors across levels interacted to shape pathways toward well-being. For example, the case presented in the tenuous recovery trajectory pointed to the ways that racism, or not having “as many chances as white people,” directly shaped feelings of worthlessness. Additionally, the presence of family conflict thwarted the participant’s capacity to reach out to family members for support with psychological vulnerabilities. Without daily interpersonal interactions—or proximal processes (Bronfenbrenner & Morris, 1998)—to offset internal struggles, the adolescent attempted suicide. Afterward, therapy helped this participant to overcome some of the strongest predictors for repeat suicidal behaviors, including hopelessness and suicidal ideation (Goldston et al., 2016; Spirito et al., 2003). Yet, because family conflict continued to characterize her everyday life after the suicide attempt, we described her pathway toward well-being as tenuous.

Taken together, our findings support an emerging empirical literature suggesting that youth can demonstrate considerable resilience following a suicide attempt (Bostik & Everall, 2007; Bergmans, Langley, Links, & Lavery, 2009; Heller, 2014). Although a history of suicidal behaviors has emerged as one of the most significant predictors of future attempts (Goldston et al., 1999; Miranda et al., 2008), there is a need for additional research that can elucidate mechanisms that contribute to resilience following a suicide attempt. Doing so can help to reveal important protective factors that might facilitate prevention and intervention efforts for suicidal youth.

For example, in our study, the majority of participants reported a reduction in feelings of hopelessness and suicidal thoughts. The cultivation of coping strategies and cognitive strengths (e.g., positive ethnic identity), coupled with positive changes in micro- and meso-level factors (e.g., family connection, new school community), was notable among girls in the resilient trajectory. Additionally, girls in the tenuous growth trajectories expressed considerable determination to overcome life circumstances that were often beyond their control, including exo- and macro-level factors, such as discrimination and criminalization of youth risk behaviors (e.g., fighting at school).

In our study, girls committed considerable time and effort toward recovery following a suicide attempt, most often in the form of therapy. Benefits from therapy, at least from the perspective of girls in our study, seem to derive from a trusted, empathetic relationship between the therapist and adolescent. Our findings compliment the results of previous studies of treatment post-suicide attempt, which support the moderating effect of a strong therapeutic alliance in reducing suicidal ideation (Bryan, Corso, Corso, Kanzler, Ray-Sannerud & Morrow, 2012; Long, Manktelow, & Tracey, 2014; Gysin-Maillart, Soravia, Gemperli &Michel, 2017).

However, it is important to note that some participants did not experience clinical benefits as a result of therapy. This finding suggests that although access to mental health treatment is important, it might not be sufficient (Jordan et al., 2012). In a study examining the lifetime prevalence of suicidal behaviors among U.S. adolescents, Nock and colleagues found that over 80% of participants had received a form of mental health treatment, and over 55% of participants had started treatment prior to the onset of suicidal behaviors (Nock et al., 2013).

These findings carry implications for interventions for Latina suicidal teens. Overall, the majority of girls in our sample benefited from clinical services despite growing up in a context of social deprivation and violence. However, for the girls who presented chronic stress trajectories, clinical services were insufficient and failed to address continuous stressors and vulnerabilities that they experienced. This suggests that for a group of suicidal Latina teens, traditional clinical interventions post-suicide attempt should be accompanied by comprehensive social services addressing risk factors across ecodevelopmental contexts (Long et al., 2014). Clinical services for suicidal adolescents should thus vary in intensity and scope depending on a teen’s baseline assessment regarding their prospective resilience.

Social support also emerged as a critical factor that could shape well-being in the aftermath of a suicide attempt. This finding resonates with the growing empirical literature supporting the Interpersonal Theory of Suicide (IPTS) (Joiner, 2005; Van Orden, Witte, Cukrowicz, Braithwaite, Selby, & Joiner, 2010). For example, lack of social support, or thwarted belongingness, was evident across participants in our sample. In contrast, the availability of social support, particularly through mental health care, was a salient protective factor among girls in the resilient and tenuous recovery trajectories. In some cases, therapy helped girls to improve their connections with the family. In those cases where family conflict continued, therapy helped adolescents to develop coping strategies to manage the conflict or perceive the conflict differently.

Importantly, some of the complexity in understanding trajectories in the aftermath of a suicide attempt cannot be adequately captured through IPTS, in large part, due to the theory’s primary focus on individual-level factors. In our study, for example, the availability of social support was shaped through macro-level processes, such as culture, immigration, ethnic discrimination, or incarceration (Mandracchia & Smith, 2015; Kene, Brabeck, Kelly & DiCicco, 2016). It is important to be mindful of the ways in which broader processes shape or constrain resilience following a suicide attempt, particularly within communities of color, such as Latina youth. In our study, for example, developing a positive ethnic identity helped to promote emotional resiliency. The literature has consistently shown that a positive ethnic identity can mitigate the potentially harmful consequences associated with culturally-relevant stressors, such as discrimination (Umaña-Taylor & Updegraff, 2007).

To date, researchers have directed little attention toward understanding Latina’s teens experiences, perceptions, and negotiations of ethnic identity and structures in the United States, and how these processes shape the onset of suicidal thoughts and behaviors (Viruell-Fuentes, 2007; Viruell-Fuentes, Miranda, & Abdulrahim, 2012). Our results highlight the importance of examining suicidal behaviors across micro-, meso-, exo-, and macro-levels, as such research is likely to provide important insights to shape the development of population-specific and ecodevelopmentally-appropriate clinical interventions for suicidal behaviors and policy advocacy strategies to address contexts of suffering to reduce statistics of suicidal behaviors among Latinas.

Limitations

There are several factors that limit the generalizability of these results. As an exploratory study with a small sample, our findings can only provide suggestions for pathways that potentially shape well-being in the aftermath of a suicide attempt. Additional research with larger samples that cross-cut geographic locale, socioeconomic status, and psychiatric diagnosis can evaluate and test the implications of our findings to develop a richer empirical base. Additionally, we only focus on the experiences of Latina girls. Although Latina adolescents have higher rates of suicide attempts than Latino adolescents (Romero et al., 2014), youth of color across the gender spectrum exhibit high risk for suicidal behaviors (CDC, 2018). Additional research that includes additional dimensions of identity would help to elucidate and clarify the ways in which intersectionality influences pathways of recovery following a suicide attempt.

Moreover, there are several important limitations stemming from the longitudinal research design. In our study, we reported a high rate of attrition. Individuals with histories of suicidal behaviors have been shown to be especially difficult to retain in longitudinal studies, and attrition rates have been reported to be as high as 90% (Boergers & Spirito, 2003). Attrition is especially pronounced in studies that collect data beyond a six-month time period (Gibbons, Stirman, Brown, & Beck, 2010). Because attrition is more likely among participants at highest risk for suicide, future studies should consider additional outreach activities to reduce attrition, including conducting interviews in home settings and collecting phone and mail contacts of extended family members (Gibbons et al., 2010).

Additionally, our study collected data at only two time points. For the purposes of a longitudinal analysis, collecting data at more frequent time intervals would have provided additional information about the girls’ experiences, resulting in a richer analysis and a more vivid conceptual understanding of their recovery over time. We also did not collect longitudinal data with additional family members. Given the importance of proximal processes to adolescent developmental outcomes (Bronfenbrenner & Morris, 1998), it is critical to gather information about how family members contribute to a girl’s recovery.

Finally, it is important to note that we conducted the study as social media technologies were just beginning to emerge (e.g., Internet chat rooms), and prior to the development of the smart phone (Luxton, June, & Fairall, 2012). Several studies have pointed to the importance of social media use as both a risk and protective factor in adolescent suicidal behaviors. For example, Becker and Schmidt found that adolescent use of social media can shape pro-suicidal behaviors by fostering cyberbullying and reinforcing self-injurious behaviors (Becker & Schmidt, 2004). Social networking sites can also support suicide prevention through increased awareness of suicide warning signs, crisis support, and enhanced social connection (Best, Manktelow, & Taylor, 2014). Although research indicates that the effects of social media on suicide risk appear to be especially pronounced in girls (Twenge, Joiner, Rogers, & Martin, 2018), studies rarely investigate racial and ethnic differences in social media use and disparities in suicide risk. Future research should consider the role of social media use and implications for suicide risk and prevention among Latina teens.

Despite these limitations, the study represents the first qualitative, longitudinal project that examines trajectories of well-being among Latina teens who have attempted suicide. Our results are consistent with the existing understanding of suicide, while extending the literature to explore the protective factors and risk factors for Latina adolescents in treatment following a suicide attempt. Our findings have important clinical and community implications and underscore the importance of prevention and intervention programs that foster adolescents’ connectivity across ecodevelopmental contexts.

Biography

Lauren E. Gulbas, PhD, is Assistant Professor in the Steve Hicks School of Social Work at the University of Texas, Austin, TX, USA.

Samantha Guz, MSSW, is a graduate student in the Steve Hicks School of Social Work at the University of Texas, Austin, TX, USA.

Carolina Hausmann-Stabile, PhD, is Assistant Professor in the Graduate School of Social Work and Social Research at Bryn Mawr College, Bryn Mawr, PA, USA.

Hannah S. Szlyk, LCSW, PhD, is a postdoctoral fellow in the George Warren Brown School of Social Work at Washington University in St Louis, St. Louis, MO, USA.

Luis H. Zayas, PhD, is Professor and Dean of the School of Social Work at the University of Texas, Austin, TX USA.