Threat sensitivity mediates the association between perceived safety and anxiety in adolescents from urban communities

Abstract

Purpose

Despite the prevalence of violence exposure (VE) among youth and its link to increased risk for anxiety and related psychiatric disorders, few studies have examined potential mediators. The present study investigated the relationships among lifetime VE, perceived safety, threat sensitivity, and anxiety symptoms in adolescents from an urban minority community (Detroit, MI).

Methods

A sample of 54 adolescents (ages 14–18, 55.6% female) self-reported on VE, safety, threat sensitivity, and anxiety symptoms in an online survey. Linear regression analyses were conducted to examine the associations among these variables.

Results

Results indicated that perceived safety, but not VE, was significantly associated with lower anxiety and threat sensitivity. Furthermore, threat sensitivity mediated the association between perceived safety and anxiety symptoms.

Discussion

The findings of the present study highlight (1) the potential protective role of perceived safety and (2) threat sensitivity as a potential target for disrupting anxiety-related pathways in adolescents with VE.

Approximately 97% of Black adolescents from urban settings have been exposed to some form of violence in their lifetime [1]. Violence exposure (VE) in adolescence is associated with psychiatric outcomes, including anxiety and depression [2,3]. However, the mechanisms and potential buffers underlying the relationship between VE and anxiety are unclear.

Several protective factors may mitigate anxiety risk in youth with VE, such as perceived safety (hereafter “safety”), or a sense of security [4]. Emerging research suggests that safety may promote mental well-being in adults and positive development in youth [5–7]. However, no studies have explored safety as a potential mediator of anxiety risk in adolescents with high VE rates.

Threat sensitivity can be defined as reactivity to potentially dangerous stimuli or situations [8]. Threat sensitivity may mediate the relationship between VE and anxiety risk (e.g., among adolescent victims of peer victimization) [9]. To date, few studies have examined the associations among VE, safety, threat sensitivity, and anxiety symptoms (hereafter “anxiety“) in urban, minority youth.

In a sample of adolescents from an urban city, we hypothesized that more VE and less safety would be individually associated with more anxiety and greater threat sensitivity, and that threat sensitivity would mediate the associations between VE/safety and anxiety.

Methods

Fifty-four adolescents, ages 14 to 18, from Detroit, Michigan were surveyed between 2021 to 2023 (Table 1). Adolescents and a parent/guardian completed a research information sheet. The local institutional review board approved all study procedures.

Table 1.

Participant demographics

Variable	Mean (SD)
Age, in years	15.2 (1.3)
Violence exposure (TIHSHa)	26.7 (5.6)
Perceived safety (TIHSHa)	14.5 (1.8)
Anxiety symptoms (SCAREDb)	27.5 (15.6)
Threat sensitivity (FSSC-R-SFc)	41.7 (7.0)
	N (%)
Age
14 years	22 (40.7%)
15 years	13 (24.1%)
16 years	7 (13.0%)
17 years	10 (18.5%)
18 years	2 (3.7%)
Gender
Female	30 (55.6%)
Male	23 (42.6%)
Other	1 (1.9%)
Race/ethnicity
Black/African American	30 (55.6%)
South Asian	13 (24.1%)
Middle Eastern/North African	5 (9.3%)
Hispanic/Latino	4 (7.4%)
White	1 (1.9%)
Biracial	1 (1.9%)
East/Southeast Asian	0 (0%)
Sexual orientation
Heterosexual	43 (79.6%)
Bisexual/pansexual	7 (13.0%)
Lesbian/gay	2 (3.8%)
Missing	2 (3.8%)
Annual household income
<$24,999	15 (27.8%)
$25,000–49,000	19 (35.2%)
$50,000–74,999	18 (33.3%)
$75,000–99,999	0 (0%)
$100,000–149,999	1 (1.9%)
>$149,999	1 (1.9%)

^aThings I Have Seen and Heard Scale

^bSCARED = Screen for Child Anxiety-Related Emotional Disorders

^cFSSC-R-SF = Fear Survey Schedule for Children-Revised, Short Form

Adolescents completed a 15–30 min online survey measuring lifetime VE, anxiety (past three months), and current threat sensitivity (Supplemental Material). Adolescent VE and safety were measured using a modified 20-item version of the Things I Have Seen and Heard (TIHSH) scale [10,11]. Sixteen items correspond to VE (Cronbach’s α = .73) and four items to safety (α = .51; Table S1). Anxiety was measured using the 41-item Screen for Child Anxiety-Related Emotional Disorders (SCARED; α = .94) [12]. A score of ≥ 25 indicates a high risk of clinical anxiety [12]. Threat sensitivity was measured using the 25-item Fear Survey Schedule for Children-Revised-Short Form (FSSC-R-SF; α = .79), which assesses five subscales of fear (Table S2) [13].

Descriptive statistics were used to characterize the sample. Hypotheses were tested using linear regressions and mediation. Exploratory analyses examined subscales of threat sensitivity. Results utilized a significance level of p < .05. Mediation was performed using the PROCESS macro (version 4.2) for SPSS (model 4) using 5,000 bootstrap samples [14]. In the mediation analysis, safety was entered as the independent variable, anxiety as the dependent, and threat sensitivity as the mediator. Mediation results were considered significant if the 95% bias-corrected confidence interval (CI) of the indirect effect did not include zero. Detailed description of analyses can be found in the Supplemental Material.

Results

Nearly all participants (98%) experienced at least one VE, and 65% of the sample reported experiencing five or more different types of VE. Additionally, 39% reported typically not feeling safe in their neighborhood and 56% demonstrated risk for clinical anxiety (Table S1, Fig S1). VE and safety were negatively correlated (r(52) = −.33, p = .015). We estimated correlations with VE, safety, threat sensitivity, anxiety, and potential confounding variables (e.g., annual household income (AHI), parent education level, parent relationship status). AHI and safety were positively correlated (r(52) = .34, p = .011), but no other variables were significantly correlated.

Safety significantly predicted anxiety (β = −.53, p < .001, F(1,53) = 20.31, R² = .28) and overall threat sensitivity (β = −.28, p = .04, F(1,53) = 4.26, R² = .08), such that adolescents who felt safer in their homes and communities reported lower anxiety and threat sensitivity. Additionally, threat sensitivity positively predicted anxiety symptoms (β = .58, p < .001, F(1,53) = 26.66, R² = .34). In contrast to results for safety, VE was not associated with anxiety (β = −.04, p = .790, F(1,53) = .07, R² = .001) or threat sensitivity (β = −.04, p = .792, F(1,53) = .07, R² = .001).

After confirming significant associations between safety/anxiety, safety/threat sensitivity, and threat sensitivity/anxiety, we tested whether threat sensitivity mediated the association between safety and anxiety (Fig 1). Threat sensitivity partially mediated the negative association between safety and anxiety (completely standardized indirect effect = −.13, bootstrap SE = .06, CI = −.26 to −.03), such that adolescents who felt less safe reported higher threat sensitivity, and higher threat sensitivity, in turn, was associated with higher anxiety symptoms. This finding survived after controlling for AHI (completely standardized indirect effect = −.14, bootstrap SE = .07, CI = −.29 to −.03).

Figure 1. Mediation model.

Threat sensitivity significantly mediated the negative association between perceived safety and anxiety symptoms in adolescents. Path coefficients represent standardized regression weights. The c path coefficient represents the total effect of safety on anxiety symptoms, while the c’ path coefficient represents the direct effect. *p < .05, **p < .01, ***p < .001.

Exploratory analyses revealed that safety (β = −.37, p = .005, F(1,53) = 8.45, R² = .14) and anxiety (β = .49, p < .001, F(1,53) = 16.62, R² = .24) significantly predicted the threat sensitivity subscale score for fear of danger and death, such that lower safety and higher anxiety symptoms were associated with higher fear of danger. An exploratory mediation analysis revealed that, specifically, the fear of danger and death subscale score partially mediated the negative relationship between safety and anxiety (completely standardized indirect effect = −.13, bootstrap SE = .06, CI = −.24 to −.02).

Discussion

Contrary to our predictions, VE did not predict anxiety or threat sensitivity among urban, minority adolescents. Safety, however, significantly predicted anxiety and threat sensitivity, and threat sensitivity mediated the association between safety and anxiety. Exploratory analyses of threat sensitivity subscales revealed that fear of danger and death explained the association between safety and anxiety. Together, this study highlights altered threat sensitivity as a potential pathway to anxiety among adolescents with high VE.

The present study centers on an under-studied sample (urban, minority, lower-income), potentially explaining why our results differ from previous studies [3,15,16]. For example, 98% of participants reported experiencing one or more forms of violence, potentially restricting our ability to detect effects of VE on anxiety or threat sensitivity due to low variability in the sample. In addition, the selected VE measure does not capture type, severity, chronicity, or age of onset of VE—which may impact outcomes. Further, VE beyond a certain threshold may be adaptive for some children in dangerous environments [17].

Our safety measure captures adolescents’ lack of safety—a response to VE—rather than just their total VE. Similarly, prior studies in PTSD patients suggest the subjective response to trauma may be a better predictor of psychiatric symptoms than number of experiences [18]. Our results contribute to the growing body of research highlighting safety’s role in buffering against psychopathology in adolescents [5,6]. Interestingly, despite high levels of VE in this sample, most adolescents (61–93%) felt safe across contexts, suggesting that safety may be a common protective factor.

Consistent with our prior findings that social threat sensitivity mediated the association between peer victimization and anxiety, threat sensitivity mediated the relationship between safety and anxiety [9]. Safety may dampen response to potential threats and, in turn, reduce anxiety. Interventions that target safety and threat sensitivity, such as family-based interventions and Attention Bias Modification Therapy, may benefit adolescents with low safety [8,19,20].

When interpreting these findings, it is important to consider the limitations of this study. The relatively small sample size limits the statistical power to detect significant effects and, therefore, our ability to generalize findings to other populations or settings. Future studies should extend this research to suburban or rural youth, or youth from different sociodemographic backgrounds. The reliance on self-reported measures may introduce bias, particularly recall bias. Using objective measures (e.g., threat detection tasks, diagnostic interviews) in future studies would provide more comprehensive findings. Additionally, the cross-sectional design of this study limits our ability to infer causality. Future studies should employ a longitudinal design to provide insights into how these variables interact over time and examine the potential protective or confounding effects of access to mental health resources and other variables.

The present study suggests that a lack of perceived safety in urban adolescents alters threat sensitivity, thus increasing anxiety risk. Future studies should evaluate the role of perceived safety and threat sensitivity as potential markers of anxiety risk and points of intervention for youth exposed to violence.

Implications and Contribution.

The present study suggests that perceived safety serves as a protective factor for anxiety in adolescents from an urban community. Furthermore, a lack of perceived safety may increase threat sensitivity, thereby heightening anxiety risk in adolescents with violence exposure. These findings may offer potential targets for disrupting anxiety-related pathways in adolescents.

Abbreviations:

VE

Violence exposure

AHI

annual household income