Do local geographic characteristics moderate the association between familial incarceration and adolescent depression and anxiety? Authors: Myriam Forster, Christopher J. Rogers, Xiao Zhang, Mikaela Rojas, Natasha DeSilva, Eric Shanazari, Jennifer B. Unger, Timothy J. Grigsby Published: 2025 ABSTRACT BackgroundChildren who have, or have had, an incarcerated parent/s are at increased risk for mental health issues. However, most research has focused on the impact of incarcerated parents and not accounted for noncustodial family members’ criminal-legal justice system involvement or how neighborhood characteristics such as area deprivation (ADI: income, education, and housing quality), environmental justice (EJI: environmental burden, social and health vulnerability), or urbanicity (RUCA) may affect these relationships. To address these gaps, we investigated the association of familial incarceration (FI: parent-caregiver, parent-caregiver + another relative, or noncustodial relatives) and depression and anxiety and the role of neighborhood characteristics in these associations.MethodsData are baseline survey responses, geocoded and linked to geospatial indices of ADI, EJI, and urbanicity of high school students enrolled in a multisite study. GLMs assessed associations between FI and students’ depression and anxiety (CESD, GAD7) adjusting for age, sex, ethnicity, state, and other family adversities (i.e., parent-caregiver substance use, mental illness, suicide intimate partner violence) and whether ADI, EJI or urbanicity moderated these relationships.ResultsThe sample (N = 1,736) was on average 15.71 (SD = 1.98) years old and evenly divided by sex. Over one third of students acknowledged FI, and approximately 50% and 75% experienced some symptoms of depression and anxiety, respectively. Having a parent only (β = 0.32, 95% CI: 0.20–0.53); parent + another relative (β = 0.27, 95% CI: 0.11–0.48); or noncustodial relative (β = 0.37, 95% CI: 0.21–0.56) was associated with more frequent symptoms of depression and anxiety [(parent only β = 0.30, 95% CI: 0.23–0.53); parent + another relative (β = 0.30, 95% CI: 0.19–0.43); noncustodial relative (β = 0.37, 95% CI: 0.23–0.53)]; These relationships were most pronounced in rural vs. urban, high vs. low environmental vulnerability, and high vs. low ADI geographic contexts.ConclusionsOur findings suggest youth whose family members are involved in the criminal-legal justice systems are more vulnerable to experiencing symptoms of depression and anxiety than their peers and that living in disadvantaged contexts can exacerbate this association. Improving the health and wellbeing of this youth population will require cross disciplinary efforts that inform the development and delivery of support services designed to meet the needs of impacted adolescents. Background The United States (US) has among the highest global incarceration rates, with over 1 million adults currently held in state prisons, 200,000 in Federal facilities, and over half a million in local jails [1]. Despite significant improvement in the ethnic-racial disparities of incarceration rates, African Americans -- the most overrepresented group in the prison population -- and Hispanics, the largest ethnic minority group in the US [2], are imprisoned at 4.9 and 2.4 times the rate of non-Hispanic Whites, respectively [3, 4]. Moreover, incarcerated populations tend to be concentrated at lower levels of education and income [5], which can perpetuate the health and wealth gap of families and children in disadvantaged communities [6, 7]. Currently, approximately 5 million children (7% of the US children) have had a parent or caregiver incarcerated at some point in their lifetime [8–13] with many more youth experiencing the incarceration of noncustodial family members. Having, or having had, an incarcerated parent can increase risk for academic difficulties [14], disciplinary action and reduced school engagement [15], and compromised mental [16–19], physical [20], and behavioral health [21, 22] -- challenges that can persist into adulthood [23–27]. Given the recent rise in adolescent depression and anxiety [28, 29] and the consequences of parental incarceration for children and communities, ongoing research that investigates the impact of family member criminal-legal justice system involvement and the influence of local contexts on youth mental health is critical. Vulnerability for early life course-onset depression or anxiety is influenced by several factors including traumatic events or conditions (such as family incarceration). Adolescents are particularly susceptible to psychosocial stressors due to the biological, neurological and physical changes that occur during this developmental period [30, 31]. Over time – and especially if youths’ attempts to limit distress are unsuccessful in the face of persistent challenges -- the “wear and tear” on the structures and physiological processes involved in mood and behavior increases risk for later life psychopathology [32–37]. For families, the economic hardship due to legal costs, housing insecurity [38–41], and loss of income [42, 43] can be extreme given that most men and women’s earnings were already at the lowest end of the national income distribution pre-incarceration [44, 45]. This financial instability can heighten preexisting vulnerabilities such as mental health and substance use issues of adults [46, 47] that further erode relationships among parents, children, and other household members. In conjunction with the economic stressors placed on families, the involuntary separation from parents, the witnessing of an arrest or sentencing, and the struggle of continuing to maintain a relationship with the incarcerated parent [48] can influence young peoples’ identity processes and interpersonal relationships [44, 45]. As described by ambiguous loss theory [49, 50] a physically absent family member who is psychologically present or a physically present family member who is psychologically absent can lead to boundary ambiguity – a reshuffling of family members’ roles and confusion as to who is in or out of the family system [51, 52]. The uncertainty surrounding the relationship with an absent family member and the structural changes in the family system that shift the responsibilities and expectations among family members often jeopardize healthy development and youths’ mental health [53, 54]. Potentially compounding the effect of family stressors on physiological processes is adolescents’ sensitivity to the broader social environment [55] such as socioeconomic conditions, urbanicity, exposure to violence, and the availability of institutional resources on development [56]. The literature demonstrating the effects of context (i.e., where youth live, go to school, and play) on adolescent health [57, 58] has highlighted the need to account for the built environment in public health and prevention research. However, to date, few studies focused on family incarceration have incorporated neighborhood factors such as degree of urbanicity, neighborhood resources, and health, social, and environmental vulnerability in school and community-based studies. Generally, data indicate that living in urban, as compared to suburban or rural areas, can compromise psychological or emotional outcomes [59–61] independent of other known risk factors such as sex, ethnicity, and socioeconomic status among school age youth [62]. Although crime, pollution or noise pollution, and social isolation can elevate risk for mental health issues in urban settings, inconsistencies [63–65] suggest that the risks of urban settings may differ across youth populations or be influenced by the presence or absence of other risk factors [65]. Similarly, meta-analyses of neighborhood health, social and environmental vulnerability (limited access to affordable housing, greenspace, or health care) suggest an overlap between levels of hazardous pollutants, little greenspace, or a disproportionate number of health risks and race and poverty in the US [66]. Neighborhood context matters but may be particularly consequential for youth living in households that also experience other risk factors such as familial incarceration. Moreover, and far less well researched, is whether the incarceration of a close relative (not just custodial relatives) may be similarly consequential for youth mental health. Integrating neighborhood characteristics in the analysis of risk and protective factors for youth mental health outcomes will allow researchers and practitioners to better understand the intersection of local context (poverty, employment, housing instability) and household adversities (familial incarceration) on adolescent mental health. To explore the role of neighborhood characteristics in the relationship between FI and youth depression and anxiety we linked participant residence to publicly available census tract level data that measures health, social, and environmental burden, neighborhood disadvantage, and degree of urbanicity. We hypothesized that youth who have or have had H1a) an incarcerated parent/caregiver H1b), incarcerated noncustodial relative, or H1c) parent/caregiver plus other incarcerated relatives would have elevated symptoms of depression and H2a-c) and anxiety after adjusting for other household stressors and demographic characteristics. We also explore whether these associations vary across indices of urbanicity, area deprivation (ADI: income, education, employment, and housing quality) and environmental vulnerability (EJI: environmental burden, social vulnerability, health vulnerability). Methods Data are baseline survey responses of high school students (N = 1,736) enrolled in the [blinded] project, a quasi-experimental study designed to evaluate the impact of a prevention program tailored to meet the needs of youth whose family members interact with the justice system. Principals of schools can request the prevention program be implemented on their campus and all high schools that offer the program (California-6 schools, Oregon-3 schools, and Georgia- 6 schools) were invited to participate in the study; 87% agreed (2 schools in California declined). Research assistants visited randomly selected classrooms at each school (4 to 5 control classes per school) and the classroom hosting the prevention program (students participating in the prevention program had been enrolled for less than three weeks, limiting any potential program benefits at baseline assessment). Students were invited to participate during in-person classroom/campus visits and approximately 95% of youth whose parents consented agreed to participate (with similar rates across states, schools and classrooms). On aggregate, 78% of parents/guardians provided consent with some differences by state; 89% in California, 77% in Oregon, and 69% in Georgia. Students took the 30-minute survey that asked about their experiences at school and in their communities, ACE, household incarceration, and included behavioral and mental health screeners and received a $15 gift card upon completion. Given that three states and 13 schools were sampling sites, students attending the same schools may have shared similar characteristics relative to those who did not. However, diagnostics revealed that intraclass correlation (ICCs) across schools for anxiety and depression were less < 0.003. Surveys were geocoded by addresses to publicly available data that links place-based exposure measures to a geographically defined administrative boundary [67] according to neighborhood of residence and to aggregate neighborhood characteristics (e.g., percentage of households below the poverty line, urbanization, greenspace, social vulnerability). All study methods were approved by [BLINDED] IRB. Measures Familial incarceration (FI) Was assessed using an adapted version of the adverse childhood experience tool [68], and asked respondents whether any of the following family members had ever gone to jail or prison. Response options include caregiver, mother, father, brother, sister, aunt, uncle, grandmother, grandfather, and other. Reponses were operationalized as none = 0, parent or caregivers = 1, parent/caregiver and other family member = 2, other family member = 3. Depression Was assessed with the 10-item Center for Epidemiologic Studies Depression Scale (CES-D) [69]. Sample questions include “During the past week:” “I was bothered by things that usually don’t bother me,” “I had trouble keeping my mind on what I was doing,” “I felt that everything I did was an effort,” “I felt fearful.” Responses were coded ‘Less than 1 day or never = 0,’ ‘1–2 days = 1,’ ‘3–4 days = 2,’ ‘5–7 days = 3.’ The depression scale was calculated by summing the responses to the 10 items and had a range from 0 to 30, with higher scores indicating more severe depressive symptoms (α = 0.80). For ease of interpretation and comparison, depression was standardized with mean = 0 and SD = 1. Anxiety Was measured with the 7-item generalized anxiety disorder GAD-7 instrument [70]. Respondents were asked: “How often have you been bothered by the following over the past 2 weeks?” with sample items including “Feeling nervous, anxious, or on edge,” “Not being able to stop or control worrying,” “Being so restless that it’s hard to sit still,” “Becoming easily annoyed or irritable,” and “Feeling afraid as if something awful might happen.” Response options were coded ‘0 = not at all,’ ‘1 = several days,’ ‘2 = more than half the days,’ and ‘3 = nearly every day.’ Anxiety items were summed and had a range from 0 to 21, with higher scores indicating more severe and frequent symptoms of anxiety (α = 0.92). Anxiety was standardized with mean = 0 and SD = 1. The Area Deprivation Index(ADI) [71, 72], originally created by the U.S. Health Resources and Services Administration, includes 17 education, employment, housing quality, and poverty measures originally drawn from long-form Census data and updated to incorporate the American Community Survey (ACS) data. Participant addresses were geocoded utilizing ArcGIS (Esri, Redlands, CA) and were matched with the Federal Information Processing Standards (FIPS) number, corresponding to United States Census Block Groups, and assigned an ADI national ranking according to the FIPS from 1 to 100, with a higher ADI rank indicating greater disadvantage. ADI was standardized with mean = 0 and SD = 1. The Environmental justice Index (EJI) [73] is a tool to quantify and rank census tracts on environmental burden, social vulnerability, and health factors. The EJI uses data from the U.S. Census Bureau, the U.S. Environmental Protection Agency, the U.S. Mine Safety and Health Administration, the U.S. Department of Transportation, OpenStreetMap, the U.S. Geospatial Survey, the U.S. Department of Agriculture, the U.S. Federal Emergency Management Agency, the National Aeronautics and Space Administration, the National Oceanic and Atmospheric Administration, and the U.S. Centers for Disease Control and Prevention to determine the cumulative impacts of environmental injustice for more than 82,000 U.S. census tracts. The base EJI ranks each tract on 36 environmental, social, and health factors and groups them into three overarching modules and ten domains. The overall EJI score is calculated by summing the ranks of three modules: the Environmental Burden Module, the Social Vulnerability Module, and the Health Vulnerability Module; each module represents an important aspect of cumulative impacts. The final EJI ranking ranges from 0 to 1 [74]. Participant addresses were geocoded and matched with the FIPS number, corresponding to the United States Census tract and assigned an EJI score according to the FIPS from 0 to 1. Higher scores indicate higher environmental vulnerability. For analysis, and because states have such vastly different environmental, social, and health vulnerability, EJI quartiles were created based on median of each state and were coded 1–4, 1 = least vulnerable, 4 = most vulnerable. Rural-Urban commuting area (RUCA) Codes, a detailed and flexible scheme for delineating sub-county components of rural and urban areas and which have been updated using data from the 2010 decennial census and the 2006-10 American Community Survey (ACS) [75]. Participant addresses were geocoded and were matched with the FIPS number, corresponding to the United States Census tract, and participants assigned with RUCA according to the FIPS from 1 to 10. Codes 1–3 are classified as urban, 4–6 as suburban and 7–10 as rural areas. Sex at birth was coded as male = 0 and female = 1. Age was calculated as a continuous variable by subtracting respondents’ date of birth from their survey completion date. Ethnicity was coded as Non-Hispanic White = 0, Non-Hispanic Black/African American = 1, Asian/Pacific Islander = 2, Hispanic = 3, and Multi-Ethnic = 4. State was coded as California = 0, Georgia = 1, and Oregon = 2. Adverse Childhood Experiences (ACE) were assessed using 6 items derived from the original ACE study [68]. Parental incarceration was excluded as we assess this association independently. The questions asked respondents before age 18, whether they lived with a caregiver/parent who was (a) mentally ill, (b) had attempted suicide, (c) was separated divorced, (d) misused illicit drugs, (e) misused alcohol, was (f) involved in intimate partner violence and operationalized as an index ranging from 0 to 6. Analyses The analytic sample was comprised of participants (N = 1,736) who provided full responses to survey items included in statistical models; students excluded for incomplete data (n = 218) were more likely to be male and live in GA. To test our hypothesized associations, depression and anxiety symptom frequency were regressed on family incarceration in linear regression models, adjusting for demographic covariates, other household adversities, and geocoded indices of environmental vulnerability, resource deprivation and urbanicity. Due to the pronounced ethnic disparities in incarceration rates, we also explored ethnic differences in the family incarceration and mental health symptomology relationship. Due to cell size of different family incarceration categories, we dichotomized family incarceration when testing for moderation by EJI, ADI and RUCA. Three separate models assessed moderation by each geographic characteristic (family incarceration*EJI quartiles; family incarceration*ADI; family incarceration*RUCA) and results of moderation by EJI quartiles, ADI, and urbanicity were converted to probabilities using the delta method with figures graphically depicting group difference. To account for multiple tests, we applied a Bonferroni adjustment with an alpha of 0.006 used to determine statistical significance; all analyses were conducted using Stata version 18 (Stata Corp., 2023). Results As seen in Table 1, the sample was on average 15.75 (SD = 1.98) years old and evenly divided by sex (males 48%, females 52%). Slightly less than one third of students identified as non-Hispanic White (31%) followed by African American (29%), Hispanic (24%), Multiethnic (9%), and Asian/Pacific Islander (7%). One in three students (33%) reported having a family member incarcerated: 9% had an incarcerated caregiver/parent, 11% an incarcerated noncustodial family member such as a sibling, aunt or uncle, and 13% reported an incarcerated caregiver/parent plus another incarcerated family member. Slightly less than half of respondents (45%) acknowledge experiencing childhood adversities other than family incarceration and 48% and 75% of students experienced at least some symptoms of depression or anxiety in the past 2-weeks, respectively. Table 1Sample characteristics (N = 1,736) by stateCalifornia Frequency (%) n = 547Georgia Frequency (%) n = 917Oregon Frequency (%) n = 272Total Frequency (%) N = 1,736State differences (p-value)Sex at birth Female301 (55%)468 (51%)136 (50%)905 (52%)Ethnicity Non-Hispanic White151 (27%)329 (36%)52 (19%)532 (31%)p <.001 African American41 (7%)384 (42%)76 (28%)501 (29%) Hispanic225 (41%)119 (13%)92 (34%)436 (25%) Multiethnic60 (11%)63 (7%)33 (12%)156 (9%) Asian Pacific-Islander71 (13%)19 (2%)19 (7%)109 (6%)FI None443 (81%)550 (60%)171 (63%)1,163 (67%)p <.01 Parent/caregiver only44 (8%)110 (12%)31 (11%)186 (11%) Parent + Other33 (6%)110 (12%)30 (11%)173 (10%) Noncustodial relatives27 (5%)147 (16%)41 (15%)215 (13%)EJI Quartile 1 (least vulnerable)175 (32%)201 (22%)36 (13%)412 (23%)p <.001 Quartile 255 (10%)83 (9%)19 (7%)157 (9%) Quartile 370 (13%)193 (21%)95 (35%)358 (20%) Quartile 4 (most vulnerable)247 (45%)440 (48%)122 (45%)809 (47%)M (SD) Depression11.24 (6.30)10.30 (6.57)11.10 (6.07)10.72 (6.43) Anxiety7.80 (5.02)6.87 (6.29)7.53 (6.01)7.27 (6.14) Age15.57 (1.32)15.16 (2.23)15.70 (1.58)15.69 (1.99) ACE1.08 (1.16)1.10 (1.15)1.23 (1.42)1.00 (1.21) ADI National12.93 (16.97)70.37 (16.14)31.79 (12.60)45.87 (30.92)p <.001 RUCA1.16 (0.69)3.91 (1.36)1.12 (0.60)2.69 (1.74)n Frequency, M Mean, SD Standard Deviation % Valid percent, FI Family Incarceration, ACE Adverse Childhood Experience, EJI Environmental Justice Index, RUCA Rural-Urban Commuting Area, ADI Area Deprivation Index Table 1Sample characteristics (N = 1,736) by stateCalifornia Frequency (%) n = 547Georgia Frequency (%) n = 917Oregon Frequency (%) n = 272Total Frequency (%) N = 1,736State differences (p-value)Sex at birth Female301 (55%)468 (51%)136 (50%)905 (52%)Ethnicity Non-Hispanic White151 (27%)329 (36%)52 (19%)532 (31%)p <.001 African American41 (7%)384 (42%)76 (28%)501 (29%) Hispanic225 (41%)119 (13%)92 (34%)436 (25%) Multiethnic60 (11%)63 (7%)33 (12%)156 (9%) Asian Pacific-Islander71 (13%)19 (2%)19 (7%)109 (6%)FI None443 (81%)550 (60%)171 (63%)1,163 (67%)p <.01 Parent/caregiver only44 (8%)110 (12%)31 (11%)186 (11%) Parent + Other33 (6%)110 (12%)30 (11%)173 (10%) Noncustodial relatives27 (5%)147 (16%)41 (15%)215 (13%)EJI Quartile 1 (least vulnerable)175 (32%)201 (22%)36 (13%)412 (23%)p <.001 Quartile 255 (10%)83 (9%)19 (7%)157 (9%) Quartile 370 (13%)193 (21%)95 (35%)358 (20%) Quartile 4 (most vulnerable)247 (45%)440 (48%)122 (45%)809 (47%)M (SD) Depression11.24 (6.30)10.30 (6.57)11.10 (6.07)10.72 (6.43) Anxiety7.80 (5.02)6.87 (6.29)7.53 (6.01)7.27 (6.14) Age15.57 (1.32)15.16 (2.23)15.70 (1.58)15.69 (1.99) ACE1.08 (1.16)1.10 (1.15)1.23 (1.42)1.00 (1.21) ADI National12.93 (16.97)70.37 (16.14)31.79 (12.60)45.87 (30.92)p <.001 RUCA1.16 (0.69)3.91 (1.36)1.12 (0.60)2.69 (1.74)n Frequency, M Mean, SD Standard Deviation % Valid percent, FI Family Incarceration, ACE Adverse Childhood Experience, EJI Environmental Justice Index, RUCA Rural-Urban Commuting Area, ADI Area Deprivation Index Sample characteristics (N = 1,736) by state n Frequency, M Mean, SD Standard Deviation % Valid percent, FI Family Incarceration, ACE Adverse Childhood Experience, EJI Environmental Justice Index, RUCA Rural-Urban Commuting Area, ADI Area Deprivation Index Based on the calculated quartiles of the environmental justice index for each state, 23% of students lived in the least vulnerable, highest quartile, followed by 9% in the second, 20% in the third and 48% in the fourth, the most vulnerable quartile. On average students had a 46.98 (SD = 20.41) score on the national 100-point area deprivation index with 100 representing the highest level of disadvantage. As to degree of urbanicity, slightly less than half (48%) of the sample lived in highly urbanized context (RUCA = 1–3), followed by approximately 35% living in suburban settings (score of 4–6) with the rest residing in varying degrees of rurality (7–10). Results of multivariable linear regression models (Table 2) suggest that any type of familial incarceration was associated with significantly more frequent symptoms of depression and anxiety, after adjusting for demographic covariates, other household adversities, and geocoded indices of environmental vulnerability, resource deprivation and urbanicity. Having an incarcerated parent-caregiver (β = 0.32, 95% CI: 0.19–0.49), noncustodial relative (β = 0.37, 95% CI: 0.07–0.38), or an incarcerated parent-caregiver plus another relative (β = 0.26, 95% CI: 0.10–0.42)) were all associated with more symptoms of depression adjusting for all covariates. Similarly, having an incarcerated parent or caregiver (β = 0.30, 95% CI: 0.23–0.53), noncustodial relatives (β = 0.37, 95% CI: 0.23–0.53), or a parent plus another relative (β = 0.30, 95% CI: 0.19–0.46) was associated with more frequent symptoms of anxiety. Table 2Results of models assessing the association between Familial incarceration and depression and anxiety symptomsVariablesDepression β (95% CI)Anxiety β (95% CI)FI None-- ref ---- ref -- Parent/caregiver only0.32 (0.19, 0.49)*0.30 (0.23, 0.53)* Other0.37 (0.07, 0.38)*0.37 (0.23, 0.53)* Parent + Other0.27 (0.10, 0.42)*0.30 (0.19, 0.46)*EJI Quartile 1 Least vulnerable-- ref ---- ref -- Quartile 2−0.01 (−0.18, 0.16)0.02 (−0.14, 0.19) Quartile 3−0.09 (−0.23, 0.16)−0.06 (−0.19, 0.08) Quartile 4 Most vulnerable−0.08 (−0.19, 0.03)−0.08 (−0.19, 0.04) Standardize ADI National0.06 (−0.03, 0.16)0.08 (−0.02, 0.17) RUCA−0.01 (−0.05, 0.03)0.01 (−0.04, 0.04)State California--ref----ref-- Oregon−0.31 (−0.56, −0.14)*−0.19 (−0.32, −0.04) Georgia−0.23 (−0.38, −0.02)−0.21 (−0.39, −0.08)*Sex Male-- ref ---- ref -- Female0.30 (0.20, 0.39)*0.42 (0.33, 0.51)* Age−0.02 (−0.04, 0.01)−0.02 (−0.45, −0.00)Ethnicity Non-Hispanic White-- ref ---- ref -- African American0.08 (−0.04, 0.20)0.01 (−0.10, 0.13) Asian/Pacific Islander−0.01 (−0.19, 0.17)0.03 (−0.14, 0.21) Hispanic−0.07 (−0.19, 0.05)0.01 (−0.12, 0.11) Multi-Ethnic0.01 (−0.15, 0.18)0.10 (−0.06, 0.26) ACE0.23 (0.18, 0.28)*0.23 (0.17, 0.27)*Moderation results1 FI (binary)0.29 (0.18, 0.39)*0.32 (0.22, 0.43)* FI*ADI0.14 (0.06, 0.29)*0.17 (0.05, 0.28)*FI*EJI Q1-- ref ---- ref -- Q2---- Q30.10 (0.01, 0.23)^0.13 (0.07, 0.34)* Q40.13 (0.01, 0.24)^0.10 (0.03, 0.29)^ FI*RUCA0.08 (0.02, 0.14)*0.09 (0.03, 0.14)*FI Family Incarceration EJI Environmental Justice Index, ADI Area Deprivation Index, RUCA Rural-Urban Commuting Area, ACE Adverse Childhood Experiences*p <.006, ^ <0.05 but not significant at the < 0.006 level1 Interaction terms were entered in separate (n = 3) models Table 2Results of models assessing the association between Familial incarceration and depression and anxiety symptomsVariablesDepression β (95% CI)Anxiety β (95% CI)FI None-- ref ---- ref -- Parent/caregiver only0.32 (0.19, 0.49)*0.30 (0.23, 0.53)* Other0.37 (0.07, 0.38)*0.37 (0.23, 0.53)* Parent + Other0.27 (0.10, 0.42)*0.30 (0.19, 0.46)*EJI Quartile 1 Least vulnerable-- ref ---- ref -- Quartile 2−0.01 (−0.18, 0.16)0.02 (−0.14, 0.19) Quartile 3−0.09 (−0.23, 0.16)−0.06 (−0.19, 0.08) Quartile 4 Most vulnerable−0.08 (−0.19, 0.03)−0.08 (−0.19, 0.04) Standardize ADI National0.06 (−0.03, 0.16)0.08 (−0.02, 0.17) RUCA−0.01 (−0.05, 0.03)0.01 (−0.04, 0.04)State California--ref----ref-- Oregon−0.31 (−0.56, −0.14)*−0.19 (−0.32, −0.04) Georgia−0.23 (−0.38, −0.02)−0.21 (−0.39, −0.08)*Sex Male-- ref ---- ref -- Female0.30 (0.20, 0.39)*0.42 (0.33, 0.51)* Age−0.02 (−0.04, 0.01)−0.02 (−0.45, −0.00)Ethnicity Non-Hispanic White-- ref ---- ref -- African American0.08 (−0.04, 0.20)0.01 (−0.10, 0.13) Asian/Pacific Islander−0.01 (−0.19, 0.17)0.03 (−0.14, 0.21) Hispanic−0.07 (−0.19, 0.05)0.01 (−0.12, 0.11) Multi-Ethnic0.01 (−0.15, 0.18)0.10 (−0.06, 0.26) ACE0.23 (0.18, 0.28)*0.23 (0.17, 0.27)*Moderation results1 FI (binary)0.29 (0.18, 0.39)*0.32 (0.22, 0.43)* FI*ADI0.14 (0.06, 0.29)*0.17 (0.05, 0.28)*FI*EJI Q1-- ref ---- ref -- Q2---- Q30.10 (0.01, 0.23)^0.13 (0.07, 0.34)* Q40.13 (0.01, 0.24)^0.10 (0.03, 0.29)^ FI*RUCA0.08 (0.02, 0.14)*0.09 (0.03, 0.14)*FI Family Incarceration EJI Environmental Justice Index, ADI Area Deprivation Index, RUCA Rural-Urban Commuting Area, ACE Adverse Childhood Experiences*p <.006, ^ <0.05 but not significant at the < 0.006 level1 Interaction terms were entered in separate (n = 3) models Results of models assessing the association between Familial incarceration and depression and anxiety symptoms FI Family Incarceration EJI Environmental Justice Index, ADI Area Deprivation Index, RUCA Rural-Urban Commuting Area, ACE Adverse Childhood Experiences *p <.006, ^ <0.05 but not significant at the < 0.006 level 1 Interaction terms were entered in separate (n = 3) models Results of individual models testing for interaction by EJI, ADI and RUCA are presented in Table 2. The relationship between familial incarceration and depression and anxiety varied across geographic characteristics although, with the exception of the lower quartiles of EJI and anxiety, moderation by EJI was significant at the α = 0.05 level (but the Bonferroni adjusted significance level of α = 006). As seen in Figs. 1a-b, in contrast to youth living in low environmental burden neighborhoods, youth with a history of family incarceration and living in neighborhoods with the highest environmental vulnerability reported more symptoms of depression and anxiety than their peers without FI. A similar trend was found for youth who report FI and who were living in neighborhoods with higher area deprivation indices. Youth with a history of FI were significantly more likely than their peers to experience symptoms of depression and anxiety in contrast to youth living in low area deprivation contexts (Fig. 2a-b). Lastly, relative to youth who report FI and live in highly urban areas, youth with FI who live outside of urban areas [76] experienced more symptoms of depression and anxiety than their peers without FI (Fig. 3 a-b). Fig. 1The association between FI and (a) depression and (b) anxiety by EJI The association between FI and (a) depression and (b) anxiety by EJI Fig. 2The association between FI and (a) depression and (b) anxiety by ADI The association between FI and (a) depression and (b) anxiety by ADI Fig. 3The association between FI and (a) depression and (b) anxiety by RUCA The association between FI and (a) depression and (b) anxiety by RUCA Discussion This is one of the few studies to explore the association between family incarceration, including parent/caregiver and noncustodial family members, and youths’ depression and anxiety, as well as variation in these associations across local neighborhood characteristics. Although national estimates suggest about 7% of the youth population has experienced the incarceration of parent before the age of 18, the percentage of youth who have or have had an incarcerated parent can vary widely by state (ranging between 3% and 13% nationally) [77]. According to the National Survey of Children’s Health (NSCH) and prior work investigating the rates of parental incarceration suggest that in California, Georgia, and Oregon approximately 8%, 8%, and 11%, of youth report parental incarceration, respectively [76]. Our results are similar to NSCH prevalence rates (slightly higher in Oregon); however, we included the incarceration of any family member, and found a substantial proportion of the youth population, up to 33% in some communities, contend with family member justice system involvement. Although our data are not representative of each states’ student populations, they do suggest that investigating the extent of family member incarceration can provide a context-specific understanding of the regional/community impact of incarceration on some households and youth. As hypothesized, over and above other family adversities, youth with incarcerated noncustodial relatives or parent-caregivers, report experiencing significantly more symptoms of depression; results that align with prior research showing that having an incarcerated parent increases risk for depression [78, 79]. Similarly, any type of family incarceration was associated with elevated symptoms of anxiety- associations that were consistent across ethnic groups after accounting for neighborhood characteristics. This finding diverges from work that has found significant ethnic differences in the association between parental incarceration and child depression [80]. This discrepancy may, in part, be due to our reliance on participant self-reported symptomology, inclusion of any family member’s incarceration, and that our analyses are limited to the statistical association between FI and mental health symptomology and do not capture the breadth of the social, developmental, and economic impact that racial disparities in policing and sentencing has on minoritized communities and youth. We also explored whether the symptoms of depression and anxiety associated with FI varied across urbanicity, environmental vulnerability, and area deprivation. Regarding urbanicity, it is theorized that youth living in an urban setting may have higher rates of mental health disorders such as depression and anxiety than their suburban or rural counterparts due to exposure to additional social stressors and elevated rates of trauma and violence [61, 81], but findings have been inconsistent [64]. In our sample, youth with no history of FI living in densely populated urban settings had similar anxiety and depression symptom profiles as their peers who do have incarcerated family members. However, in suburban and rural settings, youth with no history of FI reported fewer symptoms of depression and anxiety than their peers with FI, even after adjusting for socioeconomic status (ADI) and other family risk factors. Although beyond the scope of this study, it is plausible that youth with a history of FI living in suburban or rural settings are more isolated and stigmatized, have fewer peers with similar experiences, and face a scarcity of resources and supports in comparison to what might be available in urban centers. Thus, the advantages associated with a reduction in stressful stimuli or violence may be less beneficial for this youth population. Similarly, youth experiencing FI and living in the most structurally disadvantaged settings or neighborhoods with the highest environmental and social vulnerability were at greater risk for experiencing more frequent and severe symptoms of depression and anxiety than their peers with no history of FI. This is particularly concerning given that a significant number of incarcerated adults have experienced high levels of poverty and family adversity themselves (many have been homeless or spent time in the foster care system) [82, 83]. Upon release, many will have difficulty accessing educational and work opportunities [8] and will likely earn a wage below the federal poverty line [84, 85]. Combined, these are common pathways by which incarceration perpetuates generational trauma and exacerbates existing disparities in poverty and unemployment [86]. The challenges that families face during, and beyond the period of incarceration, such as lack of sustainable employment and housing, so critical for reintegration [84], can have downstream effects such as diminishing a young person’s hope about their own future and the opportunities available for them and their families [87]. Difficulty meeting basic needs is often coupled with limited access to social safety nets and services and having to contend with ongoing stigma and judgement [88] which affect not only the formerly incarcerated but their families and especially young people who witness these struggles. One of the most notable findings was that over one in three students was impacted by family incarceration, many of whom - the 11% of youth with incarcerated noncustodial relatives- are often overlooked in research and services. That the consequences of traumatic stressors experienced during this period of rapid and dramatic psychological, social, and neurobiological change [89–91] can significantly comprise life course health, social, and economic outcomes [92–94] points toward the critical need for programs that support families impacted by incarceration. Moving forward, data regarding family member criminal-legal justice system involvement can provide a more comprehensive understanding of the long-term impacts and inform initiatives designed to blunt the transgenerational economic and health consequences of incarceration. For example, supporting policies that reduce barriers to work force re-entry [95], restructuring punitive parole practices [96], providing mental health and substance use treatment [97] and programs that help families reestablish and build relationships post release [98] will not only reduce recidivism and help adults rebuild their lives, but strengthen families and communities. Although there has been some funding allocated to families impacted by incarceration [99] resources remain limited and the few school-based or group programs that have been implemented and evaluated, have focused on younger children and children with incarcerated parents [100–102] as opposed to the incarceration of any family member. Thus, developing and testing trauma-informed curricula designed to foster resilience and socioemotional development that empower youth to cope with their experiences as they navigate the psychological and emotional demands of this developmental period will be a positive next step in serving this population. Strengths & Limitations Our results should be considered in the context of the following limitations. First, our data are cross-sectional and do not support causal conclusions. Second, although classrooms were randomly selected at each school site, schools participating in this study were selected because they offer a prevention program tailored to youth whose family members interact with the justice system and are not representative of all communities or states impacted by incarceration. Third, due to comparisons across youth with and without family incarceration we did not include age at which family member incarceration first occurred. Fourth, while we account for other household adversities (ACE), we did not assess child maltreatment due to ethical and legal considerations. Fifth, we cannot differentiate between current or past incarceration of a family member, whether the family member went to jail or prison, or the length of incarceration. Sixth, although most students elected to participate (95%) and consent rates were relatively high for school-based studies, parental consent varied by state (89%−69%) and responses may not reflect a schools’ entire student body. Lastly, although we include students participating in the prevention program, over 90% of prevention program participants had only been enrolled in the program for three weeks when they responded to baseline surveys such that any potential benefits of program participation would be negligible. Despite these limitations our study has several strengths. First, we asked about any family member in our assessment of FI, providing compelling preliminary evidence of the impact of noncustodial family members’ incarceration along with that of parent-caregivers on adolescents’ mental health. Second, by accounting for urbanicity, area deprivation, and environmental burden we limit confounding by socioeconomic status and highlight that the interplay between family stressors and neighborhood context can provide a more comprehensive understanding of youths’ mental health outcomes. Third, although researchers have found that exposure to parental incarceration may not necessarily exert an independent influence on child health among youth experiencing other forms of family adversity, poverty, and violence exposure, our findings suggest that after controlling for other household adversities and measures of socioeconomic status FI, over and above these stressors, had a unique association with adolescent mental health. Conclusion In spite of recent advocacy and criminal justice reform initiatives, generations of families in vulnerable communities have had to navigate the complexities of arrest, incarceration, and reintegration with few supports for impacted youth. Our findings suggest the potential benefits of coordinated efforts to limit the long-term impact of incarceration on families and to design, deliver, and evaluate prevention programs and services for the substantial proportion of school age youth whose families interact with the criminal justice system.