Introduction

The United States (U.S.) has seen a surge in depression diagnoses [1] and suicidal behaviors [2] among adolescents over the past two decades. While suicide continues to be the second leading cause of death among 10–24-year olds [3], risk factors leading to suicide often go unidentified [4, 5]. The identification of suicide risk factors, including suicidal ideation and depression, both in the community and in settings such as primary care offices and emergency rooms, is key in designing interventions that prevent suicide.

The causes of the rise in depression and suicide are unclear. Social factors such as unemployment, disengagement from education, and being from a single or no parent household are commonly associated with depression in adolescents [1]. Likewise, extensive research supports the involvement of physical, psychological, and environmental factors such as familial and social influences in suicide [6]. However, less is known about the effects of the stress caused by social hierarchies on depression and suicidal ideation in adolescents, which matters in a growing environment of income inequality [7] and opportunities for social comparisons with the expansion of social media use [8].

The mechanisms by which individual and environmental multilevel factors affect health outcomes were proposed by Cohen et al. (2016) [9] in a stage model of stress and disease. This model posits that environmental demands or stressful life events and perceived stress lead to negative emotional responses, which activate both biological systems of stress and negative coping mechanisms, ultimately leading to physiological changes and disease onset and progression.

Whitehead et al. (2016) [10] pointed out three levels in which control or autonomy affects health disparities. The micro-level (personal) level, in which people in lower ranks of society experience lower actual and perceived control over their destiny, causing chronic stress that leads to poorer health outcomes and negative behavioral responses ranging from substance abuse to ineffective coping, low self-efficacy or esteem [11], and metabolic disturbances [12]; the meso (community) level, focused on the social and built environments, that can act as a chronic stressor, damaging health over time; and macro and societal level that includes the various levels of exclusion and discrimination of certain sections of society and can lead to low status and control of these groups. These levels interact, influencing health outcomes [13], which highlights the importance of considering multilevel factors associated with health outcomes.

Following these models of stress related to social rank and health outcomes, we sought to examine multilevel risk factors associated with depression and suicidal ideation in the past month in an adolescent clinical sample. These multilevel risk factors included neighborhood and school environments, perceived social status and sense of control, family history of mental illness, and in-utero and perinatal complications.

Role of Neighborhood and School Environment on Mental Health

Research on the risk factors associated with depression has generally focused on factors such as gender, exposure to stressful events, child abuse, and family history [14,15,16,17]. More recently, studies have focused on neighborhood factors [18,19,20,21] and aspects of the school environment, such as safety [22, 23], as potentially affecting health outcomes.

In schools, students who feel safe tend to exhibit lower levels of depressive symptoms [24]. Regarding neighborhood factors, perceived (rather than actual) neighborhood characteristics are more strongly associated with mental health [25]. Signs of potential danger in the neighborhood (i.e., graffiti, drug use and dealing, and violence) have been associated with poorer mental health outcomes [26], and this association has been maintained in longitudinal studies showing that perceptions of neighborhood disorder predicted symptoms of depression at a 9-month follow up interview, even after controlling for baseline depression [27]. Lack of control has been hypothesized as a factor to explain why some neighborhood factors are stressful [11].

Subjective Social Status and Perceived Sense of Control and Their Association with Mental Health

There is a well-known association between socio-economic status (SES) and health in infants [28], children, [29] and adults [30], but this association is less clear among adolescents [31, 32]. Some studies have found inverse associations between SES and global health measures, acute conditions, and health behaviors [33, 34], while others have found little evidence of SES gradients in self-rated health, acute illness, injuries, and mental health [35]. Goodman (1999) [36] found associations with certain health outcomes like self-rated health, depression, and obesity, but not with others like asthma, suicide attempts, and sexually transmitted diseases. These inconsistencies are thought to be related to the dynamic relationships between health and SES across the life span and across health outcomes, and to measurement limitations of SES during adolescence [37].

The limitations of using objective measures of social status in adolescents can be addressed by utilizing subjective social status as a measure of social rank in adolescence. In adolescents, a meta-analysis of 44 studies [37] examining the association between subjective social status and health outcomes found that higher subjective social status was associated with better health outcomes, with a similar magnitude than the findings in adults [38] and in studies examining associations between objective measures of SES and health [33], and with strongest associations with mental health outcomes, specifically depression. There is more limited evidence of a similar association between low subjective social status, and greater suicidal thoughts and behaviors both in adults [39] and adolescents [40].

In the observations of government workers in the United Kingdom with the Whitehall studies, Marmot (2004) [41] noted that a person’s perceived social status and health gradient (or one’s improved health status with higher social rank) were both associated with their degree of autonomy or sense of control as affected by their social rank and social conditions. This degree of autonomy was higher the higher social position, affecting health [41] (The Status Syndrome, pp. 46). Perceiving oneself as being in a subordinate rank and with less resources is associated with a diminished sense of control [42]. While subjective social status and sense of control are two distinct constructs, as sense of control is associated with social power whereas subjective social status is not, both are indicators of social rank. In their studies, Kraus et al. (2009) [43] concluded that perceived social status and perceived control are “related but independent constructs, with unique predictive power” (pp. 1002).

Sense of control is a heterogeneous construct and has been conceptualized as locus of control [44], learned helplessness [45], and self-efficacy [46] among others [47]. More recently, sense of control has been understood as a combination of attributional styles and self-efficacy [48], and learned helplessness and desire for control [49]. Previous studies have also revealed an association between a low sense of control and childhood depression [50, 51]. It is likely that sense of control is a combination of a personality component [52] and a more malleable component that shifts with context and age [53], the latter being more predictive of depression [54]. This dynamic calls for a better understanding of the sense of control construct and its association with depression and suicide at different age periods and settings.

The proposed association between self-control, and depression and suicidal thoughts and behaviors is based on findings proposed by the cognitive adaptation theory that posits that when confronted with life-threatening events, people adapt by adjusting their sense of control, optimism, and self-esteem to the new situation [55]. One mechanism of this adaptation involves making downward comparisons or comparing themselves with people that are in a worse situation [56]. This increased sense of control has been proven to be associated with better physical and psychological quality of life in patients with late-stage cancer [57].

Predisposing Factors Increasing Vulnerability to Depression and Suicide

There is strong evidence that a family history of depression is a risk factor for depression. In a robust 30-year-long longitudinal study, biological children of parents with depression had a twofold increase risk of major depression and suicidal ideation when compared to children of non-depressed parents. Those children who also had a grandparent with depression were at highest risk for depression but not suicide [58]. Family history of mental illness and suicide are well-known risk factors for suicide as shown in the largest case–control longitudinal study on number of suicides and suicide risk factors [59].

Adverse in-utero and perinatal conditions such as pregnancy problems and low birth weight are another individual predisposing factor for overall psychopathology [60] and suicide risk [61]. The mechanisms for these associations are thought to be related to insults in a period of development of the stress-regulation systems [61] that may have affect later development and health outcomes [60].

The aim of this study was to explore environmental stressors related to the school and neighborhood environments, measures of social rank including overall sense of control and subjective social status, and predisposing factors such as a family history of mental illness and in-utero and perinatal complications, and their associations with depression and suicidal ideation in the past month in a clinical largely urban sample of adolescents. We hypothesized that having a family history of mental illness and in-utero/perinatal complications, and poor environmental factors would be associated with depression and suicidal ideation, whereas higher perceived sense of control and subjective social status would jointly be associated to less depression and suicidal ideation.

Materials and Methods

Procedures and Participants

The data were collected from adolescents in outpatient primary care and mental health outpatient centers between February and September of 2016 in an Eastern U.S. city. The cross-sectional survey administered via paper assessed respondents’ self-reports of mental health (e.g., depression, suicidal ideation) and biopsychosocial factors (e.g., family history of mental illness, sense of control, subjective social status, school/neighborhood environments). A chart review was also conducted to obtain information related to participant’s in-utero and perinatal complications. Potential participants referred by medical staff were provided information regarding the purpose of the study and were given an option to provide assent and participate in the survey or quit the survey. Parents of assenting adolescents provided written consent. A research staff oversaw the completion of the survey in the waiting room or in a conference room provided by the clinics. This study only included adolescents (1) ages between 12 and 18, (2) being able to speak and read English and/or Spanish, and (3) being accompanied by a parent or guardian who could provide consent. Based on previous studies on sample size requirements for SEM [62], a sample of 180 participants is sufficient (estimated power > 0.80), but we aimed to recruit more to account for potential (20%) incomplete data. Of 205 potential participants, the final data analytic sample for the present study included 197 cases of adolescents who met all inclusion criteria. The study was approved by the University’s Institutional Review Board (IRB) where this study was conducted.

Measures

Outcome Variables: Depression and Suicidal Ideation in the Past Month

The participants reported their symptoms of depression during the last two weeks on the Patient Health Questionnaire-Adolescent Version (PHQ-A) [63] that used the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM–IV) diagnostic criteria to assess depressive symptomatology (i.e., low mood, anhedonia, trouble with sleep and appetite, lack of energy, trouble concentrating, psychomotor retardation, and suicidal ideation). Adequate internal consistency and validity have been identified in previous research [64], and the Cronbach’s alpha of this scale was 0.90 in the present study. For the main analysis, PHQ-A, measured with a 4-point Likert type scale (0 = not at all to 3 = nearly every day), was summed into a single variable with an overall score ranging from 0 to 27. Higher scores indicate more severe depression. For the descriptive analyses, the variable depression was dichotomized with scores 10 or above indicating the presence of moderate to severe depressive symptoms [65]. For the main analyses, to avoid overlap between the two variables related to suicidality, we removed the suicide item (“Thoughts that you would be better off dead, or of hurting yourself”) from the depression variable and used a continuous variable for depression with eight items. To evaluate recent suicidal ideation [63], we used an additional item, “Has there been a time in the past month when you have had serious thoughts about ending your life?.”

Independent Variables

The participants were asked to report a family history of mental illness (i.e., depression, anxiety, schizophrenia, bipolar disorder). In-utero and perinatal complications extracted from the medical chart were assessed by exposure to substances in-utero and pregnancy problems or birth complications.

Social Rank Factor

The 17-item self-reported Overall Sense of Control (OSOC) scale [66] measured both positive (nine items) and negative (eight items) sense of control, consisting of statements related to having control over own’s life (e.g., “I am in control of my life”) or lacking control over own’s life (e.g., “I lose control over myself”). An average of the items for positive sense of control and the reverse-coded negative items (i.e., lack of control) was calculated, where higher scores indicate a higher overall sense of control (ranged from 1.18 to 5.00). Cronbach’s alpha of this scale was 0.90 in the present study. Subjective perceptions of social stratification were assessed by the MacArthur Scale of Subjective Social Status—Youth Version [67] pictorically presenting two versions of the ladder (society and school ladders) with ten steps. The participants marked where their family was located in the social hierarchy in comparison to the general society and where they were located in their school from 1 to 10, with higher scores representing a higher perceived rank. These two variables were used as continuous variables.

Environmental Factor

Negative Neighborhood Scale [25] assessed adverse events occurred in their neighborhood in the past 6 months, including drug dealing, shooting, murders, abandoned buildings, homeless people on the street, prostitution, business closing, bad schools, and graffiti and/or vandalism (0 = none, 1 = some, 2 = a lot). A sum score was calculated, with higher scores corresponding to a more negative neighborhood environment (ranging from 0 to 19). Negative School Scale [25] measured adverse events in their school that occurred in the past 6 months, including drug dealing, shooting or knifings, teachers injured by students, school equipment damage, and anger/stress. A total score was calculated, with higher numbers corresponding to a more negative school environment (ranging from 0 to 10). Prior exposure to traumatic events was assessed by the 14 items derived from part 1 of the UCLA Reaction Index Scale [68] including traumatic events related to environmental disasters, accidents, domestic and community violence, physical and sexual abuse, and death of a loved one (0 = yes, having ever experienced the listed traumatic event, 1 = no, having not experienced it). The number of traumatic events experienced by the participant at the time of the survey was summed (ranging from 0 to 9).

Covariates

Socio-demographic measures abstracted from the medical record included age upon interview, sex at birth (0 = male, 1 = female), and self-identified race (0 = White, 1 = Black, 2 = other races or mixed race). For the data analytic purpose, each of the race categories was dummy-coded. Socioeconomic status (SES) was calculated using standardized z-scores of parental employment, education, and household income, with higher scores indicating higher family SES [69].

Data Analysis

Descriptive analyses explored the characteristics of the sample, the variables of interest, and the distribution of the study variables using SPSS software version 28. Missing data for all study variables were calculated with linear interpolation. Mplus version 7.31 [70] was used for confirmatory factor analysis (CFA) and structural equation modeling (SEM). Prior to conducting the main analysis, we checked the assumptions of SEM, none of which were found to be violated. SEM is theory driven, allowing us to identify whether a prior theoretical model could be applied to observed data by testing the relations of all variables and underlying constructs simultaneously [71]. SEM was conducted in the recommended two-step approach [72]. First, a measurement model was assessed with all relevant paths set free to vary using CFA to identify the factor structure of independent variables (i.e., family and childhood health factors, social rank factors, and environmental factors). Individual items with significant factor loadings were retained only in the final CFA to obtain a well-fitting parsimonious model [73]. Then, the hypothesized structure model (Fig. 1) included constructs validated by the measurement model was tested, wherein all hypothesized paths were estimated freely (i.e., all parameters were allowed to vary in the model to simultaneously test without any equality constraints on any parameters). We evaluated which factors were associated with depression and suicidal ideation among adolescents while adjusting for relevant sociodemographic and socioeconomic covariates.

Fig. 1
figure 1

Conceptual model that hypothesizes the effects of family and childhood health, social rank, and environmental factors on depression and suicidal ideation. Plus ( +) and minus ( −) signs indicate paths with positive and negative associations, respectively

Full size image

We hypothesized that having a family history of mental illness and in-utero/perinatal complications, and poor environmental factors would be associated with an increase in the risk of depression and suicidal ideation, whereas greater social rank factors would be associated with a decrease of depression and suicidal ideation. The final model was re-specified from the hypothesized model based on prior literature and modification indices (MIs) [71]. All SEM analyses were conducted using weighted least squares mean and variance adjusted (WLSMV) estimator due to categorical observed variables (e.g., binary or ordinal). Two standardization options were used simultaneously to obtain standardized parameter estimates and standard errors of continuous (STDYX) and binary (STDY) covariates [74]. Goodness of fit was assessed by multiple-fit indices [71, 75]: chi-square (χ2) goodness-of-fit index, the Comparative Fit Index (CFI) and the Tucker–Lewis index (TLI) ≥ 0.95; the Root Mean Square Error of Approximation (RMSEA) ≤ 0.06; Weighted Root Mean Square Residual (WRMR) < 1.0; Standardized Root Mean Square Residual (SRMR) < 0.08. Path coefficients less than 0.1 indicate a small effect, those around 0.3 a medium effect, and those greater than 0.5 a large effect [71].

Results

Descriptive Statistics

Table 1 presents descriptive statistics with whole sample (N = 197). The average age of the respondents was 14.6 years (SD = 1.5). A majority of the respondents were female (63%) and Black (63%) and reported a family history of mental illness (68%) and in-utero and perinatal complications (52%). More than one-third of the sample met the criteria for moderately severe to severe depression (39%) and reported suicidal ideation (33%). Table 2 shows the bivariate relationships of moderate/severe depression and suicidal ideation in the past month by sociodemographic, family and in-utero and perinatal health, perceived control and status, social rank, and environmental variables. Adolescents with a family history of mental illness and lower perceived sense of control were more likely to report moderate or severe depression and suicidal ideation in the past month than those with no family history of mental illness or higher perceived sense of control (p < 0.05). Table 3 displays the correlations between all variables of interest.

Table 1 Sample characteristics (N = 197)
Full size table
Table 2 Depression and suicidal ideation in the past month by sociodemographic and biopsychosocial factors
Full size table
Table 3 Correlations for study variables
Full size table

Measurement Model

An initial three-factor CFA model (family and childhood health, social rank, and environmental factors) demonstrated good fit (χ2 (17) = 17.31, p = 0.433; CFI = 1.00, TLI = 1.00, RMSEA = 0.01 [90% CI 0.00–0.07, p = 0.84], WRMR = 0.58); however, a latent variable of the family and childhood health factor had only two observed variables (i.e., family history of mental illness, in-utero and perinatal complications), indicating that those two indicators would not be appropriate to be included as a latent variable and hence the posited three-factor CFA model did not appear reasonable. Based on these findings, we decided to use each of the sub-items of the family history of mental illness and in-utero/perinatal complications as a separate observed variable. The revised two-factor CFA model with social rank and environmental factors yielded an excellent fit (χ2 (7) = 7.71, p = 0.359; CFI = 0.99, TLI = 0.98, RMSEA = 0.02 [90% CI 0.01–0.09, p = 0.67], SRMR = 0.03). All standardized factor loadings were statistically significant (ranging from 0.3 to 0.8, p < 0.01) with the anticipated directions. Thus, this revised measurement model appeared reasonable and was adopted for the present study.

Structural Model

SEM was conducted to assess the effects of family history of mental illness and in-utero/perinatal complications variables, social rank factors, and environmental factors on depression and suicidal ideation in the past month, while adjusting for all relevant covariates. The full structural model demonstrated a good fit to the data (χ2 (44) = 49.69, p = 0.257; CFI = 0.97, TLI = 0.94, RMSEA = 0.03 [90% CI 0.00–0.06, p = 0.89], WRMR = 0.62). As shown in Fig. 2, the social rank factor was negatively associated with both depression (β =  − 0.58, p < 0.000) and suicidal ideation in the past month (β =  − 0.53, p < 0.001), indicating that high perceived social rank was associated with a decrease risk of depression and suicidal ideation in the past month. A significant direct effect of the environmental factor on depression (β = 0.26, p < 0.05), and suicidal ideation in the past month was found (β = 0.42, p < 0.01), indicating that negative environmental factors were associated with an increased risk of depression and suicidal ideation in the past month. Family history of mental illness had a significant positive association with depression (β = 0.29, p < 0.01) and suicidal ideation in the past month (β = 0.25, p < 0.05). In-utero and perinatal complications were positively associated with depression (β = 0.19, p < 0.05), but not with suicidal ideation in the past month (Table 4). No correlation was found between a latent variable of social rank and environmental factors in this sample.

Fig. 2
figure 2

Full structural model that assessed the effects of family and childhood health variables (family history of mental illness, in-utero/perinatal complications), social rank factors, and environmental factors on depression and suicidal ideation. Ovals present the latent variables and rectangles present the observed variables. Only significant coefficients are presented for display purposes. The standardized coefficients are presented. *p < 0.05, **p < 0.01, ***p < 0.001

Full size image
Table 4 Parameter estimates of depression and suicidal ideation in the past month
Full size table

Discussion

Following Cohen’s stage model [9] that attempts to explain the effects of stress on disease by integrating individuals’ experiences, perceptions, and physiological circumstances, the present study examined the relationships among family mental health history and in-utero/perinatal complications, social rank factor, environmental factor, and depression and suicidal ideation in the past month in a clinical sample of adolescents. The majority of participants in this sample were Black and female. Two factors (latent variables) were confirmed based on the measurement model: a social rank factor and an environmental risk related factor.

The current study explored the mechanisms of social rank on depression and suicidal ideation in the past month. Measures of sense of control and subjective social status merged into a single social rank factor that included variables related to subjective social status of the adolescent in their school, the adolescent’s family in the larger society, and the adolescent’s sense of control. In this study, greater percevied social rank was significantly and inversely associated with depression and suicidal ideation in the past month. Sense of control is associated with social rank position [43, 67] and health [42]. Prior research has also shown a significant association between low sense of control and depressive symptoms, with one fourth to half of the variation in depressive symptoms associated with family wealth being accounted for by low sense of control [76, 77]. Furthermore, a higher sense of control can be protective of depression [76] and may serve as a buffer for socio-economic status risk [78, 79].

Status indices such as the person’s sense of control and the placement in community rank hierarchy are associated with higher levels of multisystem physiological dysregulation (including the cardiovascular, endocrine, and autonomic nervous systems) [12]. There are also age differences in the connection between social ranking and allostatic load in adults, with stronger effects in younger than older adults [12]. This finding emphasizes the need to include individual- and community-level interventions related to social rank, notwithstanding the need to address macro factors of social rank [43].

A second factor, the environmental risk factor was significantly associated with depression and suicidal ideation in the past month. The association between the environmental factor that included traumatic experiences and exposure to violence in the school and the community, and suicidal ideation in the past month was consistent with prior research showing increased suicidal ideation [80] and attempts [81,82,83] in individuals with a history of adverse events in childhood [83, 84], especially violence-related events [85] and sexual and physical abuse [86,87,88]. In this study, the effects of the environment were weaker than those of the social rank factor. This finding could be due to the cumulative effects of chronic environmental stress having more long- and less short-term effects on individuals. Additionally, adolescents who lived in relatively safe homes as children, become more exposed to larger and potentially more violent environments in larger schools and neighborhoods as they grow [89, 90].

The U.S. is experiencing historically high levels of income inequality [7], which may heighten the subjective experience of social class as powerful in predicting social outcomes. In this climate, understanding stress related to social rank and to deprivation and violence in the school and neighborhood environment is important to design appropriate interventions that can buffer the effects of inequality. An encouraging finding in our study was that psychological factors related to social rank (sense of control and status) appeared to have significant weight on depression and suicidal ideation even in adverse environments, and these more malleable risk factors could be addressed in therapy. Interventions to build resilience through adolescence, such as parenting interventions, promotion of early detection of stress-related disorders, and self-help for mood and anxiety disorders (e.g., through digital apps) [91], as well as the delivery of psychological therapies by non-specialists in low-resource settings [92] could all contribute to improvements in psychological distress related to social rank. Therapeutic interventions could focus on the social comparisons leading to feelings of worthlessness [93, 94]. Third-wave psychological therapies that include components of self-validation to counter social defeat and worthlessness associated with depression and suicidality [95] are now being adapted to extreme poverty settings [96]. These initiatives need to be accompanied by efforts to reduce social inequalities in our communities, as we know that macro and structural factors such as access to education and income inequality are the strongest determinants of adolescent health [97].

Despite its strenghts, this study also presents several limitations related to methodology. It is a cross-sectional study, and as such, causal relationships among the variables included cannot be drawn. Longitudinal studies looking at long-term effects and variation of one’s subjective social status, sense of control, and school and neighborhood environments would help determine causal mechanisms. Additionally, in-utero and perinatal complications were determined as part of a chart review. These complications could have been underreported, and the data collected may have been limited by recall bias. The rest of the measures were self-reported in the survey, which may increase the risk of social desirability. Additionally, the data were oversampled from a Black and female population, which limits generalizability. While this was a relatively small sample, many studies conducted previously have not had a sufficient sample size of minority youth. There could have been differences in reporting of symptoms in this population. Furthermore, a limitation to the study of suicidal ideation was the use of a single item. Finally, the chances of a bidirectional relationship between measures of perceived social rank with depression and suicidal ideation may be higher than for the environmental risk variables and account for the stronger association. In other words, people who are depressed may perceive themselves to be of a lower social rank, and those who see themselves as lower in the social hierarchies may also tend to feel more depressed. While the same could be said about the environmental variables, these questions elicited more objectivity as they are not focused on the respondent but on the environmental factors, and the bidirectional effect may have been lower, appearing as a weaker association in the model.

Another important finding was that stress factors related to one’s environment had a stronger association to suicidal thoughts in the past month than to depression, and that perceived social rank had a strong association with both depression and suicidal ideation in the past month. Although evidence supports an association between depression and suicide, the best fitting model did not reveal a correlation between the level of depressive symptoms and suicidal ideation in the past month. This finding is key in clinical settings as it stresses the need to assess and address social rank perceptions and environmental stressors in clinical interviews beyond the focus on depressive symptoms. Further, these measures may have a role as less stigmatizing proxy for suicide risk.

In conclusion, perceptions of social rank were associated to depression and suicidal ideation and may be worthy of exploration in future longitudinal studies. In our study, these social rank factors were more strongly associated with depression and suicidal ideation than environmental risk factors related to the school and the neighborhood and family mental health history and in-utero/perinatal complications. Individual and environmental interventions that give adolescents a greater sense of control and status may be beneficial in treating depression and suicide risk.

Summary

There has been a surge in depression diagnoses and suicide among adolescents in the U.S. over the past two decades. Previous research highlights physical, psychological, and environmental factors such as familial and social influences as implicated in depression and suicide. However, less is known about the effects of the stress caused by social hierarchies on depression and suicide in adolescents, which matters in a growing environment of income inequality and opportunities for social comparisons with the expansion of social media use. Following models of stress-related to social rank and health outcomes, we sought to examine multilevel risk factors associated with depression and suicidal ideation in a clinical sample of adolescents. We hypothesized that family history of mental health, in-utero and perinatal complications, and adverse environmental factors related to the neightbodhood and school would be associated with an increase in the risk of depression and suicidal ideation in the past month, whereas greater perceived control and higher subjective social status would be associated with a decrease in depression and suicidal ideation.

Data collected in outpatient primary care and mental health outpatient centers in an Eastern U.S. city between February and September of 2016 using a cross-sectional survey assessed respondents’ self-reports of mental health (e.g., depression, suicidal ideation in the past month) and biopsychosocial factors (e.g., family history of mental illness, sense of control, subjective social status, and school/neighborhood environments). A chart review was also conducted to obtain information related to in-utero and perinatal complications. The final data analytic sample included 197 cases of adolescents who were primarily Black and female.

Two factors (latent variables) were confirmed based on the measurement model: a social rank factor and an environmental risk factor. Our measures of sense of control and subjective social status merged into one single factor of social rank, which included variables related to subjective social status of the adolescent in their school, the adolescent’s family in the larger society, and the adolescent’s sense of control. A greater social rank factor was significantly associated with a decreased risk of depression and suicidal ideation, consistent with prior research showing an association between low sense of control and depressive symptoms. We conclude that perceptions of social rank are linked to depression and suicidal ideation. In our sample, perceived control and social status factors were more strongly associated with depression and suicidal ideation than environmental risk factors related to the school and the neighborhood environment.