Skip to main content
American Journal of Epidemiology logoLink to American Journal of Epidemiology
. 2023 Nov 3;193(3):489–499. doi: 10.1093/aje/kwad214

Suicide Mortality Among Formerly Incarcerated People Compared With the General Population in North Carolina, 2000–2020

Kate Vinita Fitch ✉, Brian W Pence, David L Rosen, Vanessa E Miller, Bradley N Gaynes, Monica E Swilley-Martinez, Andrew L Kavee, Timothy S Carey, Scott K Proescholdbell, Shabbar I Ranapurwala
PMCID: PMC11484614  PMID: 37939151

Abstract

We aimed to compare rates and characteristics of suicide mortality in formerly incarcerated people with those of the general population in North Carolina. We conducted a retrospective cohort study of 266,400 people released from North Carolina state prisons between January 1, 2000, and March 1, 2020. Using direct and indirect standardization by age, sex, and calendar year, we calculated standardized suicide mortality rates and standardized mortality ratios comparing formerly incarcerated people with the North Carolina general population. We evaluated effect modification by race/ethnicity, sex, age, and firearm involvement. Formerly incarcerated people had approximately twice the overall suicide mortality of the general population for 3 years after release, with the highest rate of suicide mortality being observed in the 2-week period after release. In contrast to patterns in the general population, formerly incarcerated people had higher rates of non–firearm-involved suicide mortality than firearm-involved suicide mortality. Formerly incarcerated female, White and Hispanic/Latino, and emerging adult people had a greater elevation of suicide mortality than their general-population peers compared with other groups. These findings suggest a need for long-term support for formerly incarcerated people as they return to community living and a need to identify opportunities for interventions that reduce the harms of incarceration for especially vulnerable groups.

This article is part of a Special Collection on Mental Health.

Keywords: incarceration, mortality, prison, race/ethnicity, suicide

Abbreviations

CI

confidence interval

DAC

Department of Adult Corrections

ICD-10

International Classification of Diseases, Tenth Revision

PY

person-years

SMR

standardized mortality ratio

Suicide is a major cause of mortality in the United States, claiming 45,979 lives at an age-adjusted rate of 13.5 per 100,000 population in 2020 (1). Suicide rates in North Carolina, the state in which this research was conducted, are very similar, with a 2020 age-adjusted rate of 13.2 deaths per 100,000 population (1). North Carolina suicide rates vary by sex and race/ethnicity, with White people and males representing the highest rates of suicide relative to other groups (2).

Formerly incarcerated people in the United States and elsewhere have been shown to have an elevated risk of suicide mortality following release from incarceration (3–11). In one systematic review, Zlodre and Fazel (8) reported a range of overall standardized mortality ratios (SMRs) for postrelease suicide deaths (observed numbers:expected numbers) of 3.2 (95% confidence interval (CI): 2.5, 4.7) to 9.2 (95% CI: 5.8, 14.7) (8). When the researchers were able to produce estimates stratified by sex and race/ethnicity, they reported higher SMRs in females and White people (8). In studies of formerly incarcerated people in North Carolina specifically, an earlier study from 1980–2005 (4) found elevated suicide mortality rates for males after release from incarceration. This effect was variable by race/ethnicity, with SMRs of 2.59 (95% CI: 2.39, 2.81) and 1.16 (95% CI: 0.98, 1.36) for White and Black males, respectively (4). Another North Carolina study from 2006–2007 showed a more extreme finding, with formerly incarcerated people having approximately 4 times the suicide mortality rate of the general population (5).

Due to the age of the data in these studies, it is not known whether the elevated burden of suicide mortality in formerly incarcerated people has changed in recent years, as both the North Carolina Department of Adult Corrections (DAC) and the criminal legal system in general have changed. Some notable changes in North Carolina prison policy and practice since these studies were published include an overall decline in the state incarceration rate (12), enhanced penalties for assaultive behavior and contraband possession by incarcerated people, increases in staffing levels (13), and implementation of mandatory postrelease community supervision for all releasees with felony convictions (14), all of which could impact postrelease suicide mortality. Further, prior studies had limited sample sizes and follow-up periods, resulting in absent or imprecise stratified estimates by race and ethnicity, age, and sex, which limits identification of and intervention for the most vulnerable individuals. The absence of racial and ethnic stratification in existing literature is a large gap, given the well-documented structural racism embedded in the US criminal legal system (15–18) and links between incarceration and health (4, 19–23). Prior investigators also have not reported suicide mortality by firearm involvement, which leaves the effect of firearm ownership restrictions (24, 25) for this population on suicide mortality unexplored. Finally, prior studies did not examine how suicide mortality risk changes over time since release; given that overdose mortality is extremely high immediately after release and then declines with time (21, 22), there is potential to observe similar high-risk periods for suicide mortality which would inform intervention strategies.

The present study addresses these substantial gaps in prior research using what is, to our knowledge, the largest ongoing cohort study of formerly incarcerated people across the world (26). We updated prior postrelease suicide mortality estimates using data collected up to March 2020; examined variations in suicide mortality by time since release; assessed the role of firearm involvement in these deaths; and evaluated effect-measure modification by race and ethnicity, age, and sex. Through these analyses, we aim to present an updated, multidimensional, and thorough description of suicide mortality patterns among people released from incarceration as compared with their general-population peers.

METHODS

We conducted a retrospective cohort study of individuals released from North Carolina prisons between January 1, 2000, and March 1, 2020, to examine the rate of suicide mortality in this population as compared with the North Carolina general population. First, we calculated suicide mortality rates among formerly incarcerated people per 100,000 person-years (PY) for 2-week, 4-week, 6-month, 1-year, 2-year, and 3-year postrelease observation periods, adjusting for age, sex, and calendar year of release. Second, we calculated suicide mortality rates per 100,000 PY for the North Carolina general population, again adjusting for age, sex, and calendar year of release. Third, we estimated the SMRs comparing formerly incarcerated people with the general population. Finally, we stratified these SMRs by race/ethnicity, sex, age group, and calendar year of release to evaluate effect modification.

Data sources

We used data from the North Carolina DAC to construct the total formerly incarcerated population. DAC data included name, date of birth, sex, race, ethnicity, dates of admission and release, number of prior incarcerations, and personal identification numbers for each released person. DAC data were linked to death records from the Vital Records Unit of the North Carolina Division of Public Health using date of birth, sex, first and last name (or Soundex equivalents), and middle initial from January 1, 2000, to March 15, 2020, to identify formerly incarcerated people who had died and to calculate suicide mortality counts. The North Carolina death records also included data on North Carolina resident decedent race, ethnicity, date of death, and cause of death as determined by International Classification of Diseases, Tenth Revision (ICD-10) codes. All adult North Carolina death certificates, including those linked to DAC records, were used to calculate general population suicide mortality counts. Finally, the total North Carolina general population was constructed using 2000–2020 bridged-race population estimates from the National Center for Health Statistics (27, 28).

Participants

Eligible individuals from the formerly incarcerated population included those released from North Carolina prisons between January 1, 2000, and March 1, 2020, who were aged 18–69 years at the time of release and had any observed postrelease person-time during the study period. Numbers of deaths and person-time beyond age 69 years were excluded because of the small number of individuals released at age 70 years or older. We hypothesized that suicide mortality trends might be affected by the coronavirus disease 2019 (COVID-19) pandemic, but we did not have sufficient follow-up data to analyze postpandemic trends. Therefore, we selected a March 1, 2020, end date to permit at least 2 weeks of postrelease observation time prior to the onset of the coronavirus disease 2019 pandemic on March 15, 2020. Individuals who were released and subsequently reincarcerated accrued person-time only during the periods in which they were not incarcerated (21). For example, an individual who was released in 2012 and lived in the community for 3 months before reincarceration would contribute 2 weeks of person-time for the 2-week rate, 4 weeks of person-time for the 4-week rate, and 3 months of person-time for the 6-month rate in 2012. They would contribute no additional person-time until being released from their second incarceration. Each North Carolina general population resident aged 18–69 years counted from the bridged-race population estimates contributed 1 PY per calendar year except the year 2020, during which they contributed Inline graphic PY. Because we had only annual population estimates for the North Carolina general population, the person-time denominator was constant for the entire year and the 2-week, 4-week, 6-month, and 1-year rates for the general population were constant.

Outcomes

We identified suicide deaths using ICD-10 underlying-cause-of-death codes X60–X84, inclusive of intentional self-harm by a variety of means, including poisoning, firearm discharge, hanging, drowning, and jumping; ICD-10 code Y87.0, sequelae of intentional self-harm; and ICD-10 code U03, suicide due to terrorism. When the underlying cause of death was X72 (handgun discharge), X73 (rifle/shotgun/larger firearm discharge), or X74 (other and unspecified firearm discharge), the death was classified as a firearm suicide death; otherwise, it was classified as a nonfirearm suicide death.

Covariates

We categorized continuous age into age groups of 18–21, 22–29, 30–39, 40–49, 50–59, and 60–69 years, to provide the best balance of capturing changes in suicide mortality risk by age group, maintaining age ranges of similar size, and providing sufficient observation counts. Releasee, decedent, and North Carolina general-population sex variables were binary (male and female).

Race and ethnicity were defined differently in the formerly incarcerated and North Carolina general populations (see Web Appendix 1, available at https://doi.org/10.1093/aje/kwad214). We constructed a composite race/ethnicity variable with the categories of non-Hispanic White (hereafter called White), non-Hispanic Black (Black), non-Hispanic American Indian/Alaska Native (Native American), non-Hispanic Asian/Pacific Islander (Asian), and Hispanic/Latino of any race (Hispanic/Latino). Individuals of multiple, unknown, or another race are represented in DAC and death certificate data but not in bridged-race population estimates, so we excluded these individuals from race/ethnicity-stratified analysis. Given that racial/ethnic misclassification in death certificates is a known source of bias (29, 30), we also compared the DAC and death certificate race/ethnicity data for deceased formerly incarcerated people (Web Appendix 2).

Statistical analysis

We calculated suicide mortality rates per 100,000 PY for formerly incarcerated people by dividing the number of suicide deaths by the number of PY contributed at 2 weeks, 4 weeks, 6 months, 1 year, 2 years, and 3 years after release. We calculated suicide mortality rates per 100,000 PY for the North Carolina general population by dividing suicide deaths from North Carolina death records by the PY contributed by the annual population estimates. Using direct standardization by age, sex, and calendar year, we computed standardized suicide mortality rates for formerly incarcerated people for comparison with the general population. Then, using indirect standardization by age, sex, and calendar year, we calculated the expected number of suicide deaths among formerly incarcerated people and compared them with the observed number of deaths to estimate overall, firearm-involved, and non–firearm-involved suicide SMRs and 95% CIs. Formerly incarcerated people who survived for more than 1 year after release were aged after each year and their person-time was compared with the general population for the calendar year to which they had survived (i.e., someone released in 2011 was compared with the general population in 2012 after surviving a full year).

To examine effect modification by race/ethnicity, sex, and age group, we calculated stratified SMRs. We chose the 3-year postrelease period to avoid violations of positivity when we evaluated effect modification by race/ethnicity and sex (31). However, to evaluate effect modification by age group, we chose the 1-year postrelease period so that an individual’s age group would remain stable for the follow-up period. Finally, we stratified the 1-year suicide mortality rates and SMRs by calendar year to evaluate changes in suicide mortality over time.

RESULTS

During the study period, 266,400 people were released from North Carolina state prisons a total of 474,892 times. Of these releases, 21,231 (4.5%) resulted in death prior to the end of the study period and 241,326 (50.8%) resulted in reincarceration. A total of 2,694,140 PY were accrued, with a mean of 5.7 years and a median of 3.6 years accrued per release. In the general population, a total of 577,126 adults aged 18–69 years died during the study period.

Formerly incarcerated people were 85.7% male and had a median age of 32 years at the time of release. The majority were Black (50.4%) or White (41.6%); 5.4% were Hispanic/Latino, 1.8% were Native American, 0.3% were Asian, and 0.7% were of unknown or multiple races/ethnicities or had no race/ethnicity listed. Compared with the midstudy (2010) North Carolina general population, Black and Native American people were overrepresented in the formerly incarcerated population while White, Hispanic/Latino, and Asian people were underrepresented (Table 1).

Table 1.

Demographic Composition of the Formerly Incarcerated Population and the General Population in North Carolina, January 1, 2000–March 15, 2020

Population
Formerly Incarcerated People  
(n = 266,400)
NC General Population a  
(n = 6,456,407)
Demographic Characteristic No. % No. %
Race/ethnicity
 Black 134,175 50.4 1,387,407 21.5
 White 110,704 41.6 4,345,390 67.3
 Hispanic/Latino 14,292 5.4 485,696 7.5
 Native American 4,810 1.8 77,612 1.2
 Asian 694 0.3 160,302 2.5
 Unknown/multiple/none listed 1,725 0.7 0 0.0
 Sex
 Male 228,300 85.7 3,164,659 49.0
 Female 38,100 14.3 3,291,748 51.0
 Age group, yearsb
 18–21 46,819 9.9 563,366 8.7
 22–29 146,715 30.9 1,020,134 15.8
 30–39 145,118 30.6 1,277,374 19.8
 40–49 95,045 20.0 1,368,638 21.2
 50–59 35,051 7.4 1,277,593 19.8
 60–69 6,144 1.3 949,302 14.7

Abbreviation: NC, North Carolina.

a 2010 population aged 18–69 years.

b For formerly incarcerated people, this is age at release per 474,892 releases rather than per 266,400 individuals.

Of the 21,231 releases that resulted in death prior to the end of the study period, 889 (4.2%) were due to suicide. Of 577,126 deaths in the general population, 20,454 (3.5%) were due to suicide. In the formerly incarcerated population, suicide death was less often completed via firearm as compared with the general population, as indicated by ICD-10 cause-of-death codes. Among formerly incarcerated people, 355 (39.9%) suicide deaths involved a firearm, while in the general population, 11,507 (56.3%) suicide deaths involved a firearm. Suicide death by hanging, strangulation, and suffocation were much more common among formerly incarcerated people than in the general population, accounting for 349 (39.3%) and 4,013 (19.6%) suicide deaths, respectively (Web Table 1).

Standardized suicide mortality rates for formerly incarcerated people were highest in the 2-week postrelease period, at 37.0 per 100,000 PY. The standardized rate declined by the end of the 4-week postrelease period to 26.8 per 100,000 PY, and then climbed steadily to another peak of 33.5 per 100,000 PY at 3 years postrelease (Table 2). The SMR comparing observed suicide deaths with expected suicide deaths among formerly incarcerated people was highest during the 2-week postrelease period, at 2.2 (95% CI: 0.8, 3.7). Formerly incarcerated people had almost twice the overall suicide mortality as the general population for all time points through 3 years postrelease (Table 2).

Table 2.

Adjusted Postrelease Suicide Mortality Rates and Standardized Mortality Ratios in Formerly Incarcerated People and the General Population, by Firearm Involvement, Postrelease Observation Period, and Demographic Strata, North Carolina, January 1, 2000–March 15, 2020

Overall Suicide Firearm-Involved Suicide Non–Firearm-Involved Suicide
Mortality Rate Mortality Ratio Mortality Rate Mortality Ratio Mortality Rate Mortality Ratio
Period or Stratum NC General Population a Formerly Incarcerated Persons b SMR 95% CI NC General Population Formerly Incarcerated Persons SMR 95% CI NC General Population Formerly Incarcerated Persons SMR 95% CI
Postrelease period
 2 weeks 16.5 37.0 2.2 0.8, 3.7 9.4 12.2 1.3 0.0, 2.8 7.1 25.0 3.5 0.7, 6.4
 4 weeks 16.5 26.8 1.6 0.7, 2.5 9.4 8.2 0.9 0.0, 1.7 7.1 18.8 2.7 0.9, 4.4
 6 months 16.5 30.6 1.9 1.5, 2.2 9.4 10.9 1.2 0.8, 1.5 7.1 19.9 2.8 2.1, 3.5
 1 year 16.5 32.0 1.9 1.7, 2.2 9.4 10.8 1.1 0.9, 1.4 7.1 21.3 3.0 2.5, 3.6
 2 years 16.7 32.1 1.9 1.7, 2.1 9.5 11.7 1.2 1.0, 1.4 7.2 20.5 2.8 2.5, 3.2
 3 years 17.0 33.5 2.0 1.8, 2.2 9.7 13.4 1.4 1.2, 1.6 7.3 20.2 2.8 2.4, 3.1
Demographic stratum
 Race/ethnicityc
  White 21.4 64.8 3.0 2.7, 3.3 12.2 22.5 1.8 1.5, 2.2 9.2 42.1 4.6 4.0, 5.2
  Black 7.5 10.1 1.3 1.1, 1.6 4.3 5.5 1.3 0.9, 1.6 3.2 4.7 1.5 1.0, 1.9
  Hispanic/Latino 6.9 21.0 3.1 1.3, 4.8 3.2 8.2 2.6 0.3, 4.8 3.7 13.0 3.5 0.9, 6.1
  Native American 15.6 18.5 1.2 0.1, 2.2 7.8 7.5 1.0 0.0, 2.3 7.8 10.9 1.4 0.0, 3.0
  Asian 9.4 84.5 9.0 0.0, 21.4 3.4 46.5 13.6 0.0, 40.2 6.0 40.2 6.7 0.0, 19.8
 Sexd
  Male 24.9 47.7 1.9 1.7, 2.1 15.4 20.7 1.3 1.1, 1.6 9.5 25.7 2.7 2.4, 3.0
  Female 8.0 25.8 3.2 2.1, 4.4 3.1 7.0 2.3 0.8, 3.8 4.9 18.8 3.8 2.3, 5.4
 Age group, yearse
  18–21 11.4 35.5 3.1 1.8, 4.4 6.3 14.8 2.4 0.9, 3.8 5.2 21.7 4.2 1.8, 6.6
  22–29 14.4 30.0 2.1 1.5, 2.6 8.1 11.3 1.4 0.8, 2.0 6.3 19.3 3.1 2.0, 4.1
  30–39 15.3 29.6 1.9 1.4, 2.5 7.9 9.4 1.2 0.6, 1.7 7.3 21.1 2.9 1.9, 3.8
  40–49 18.2 32.8 1.8 1.2, 2.4 9.4 7.2 0.8 0.3, 1.3 8.8 27.2 3.1 2.0, 4.2
  50–59 19.0 13.2 0.7 0.1, 1.3 10.6 2.0 0.2 0.0, 0.6 8.0 12.8 1.6 0.2, 3.0
  60–69 15.3 46.7 3.1 0.1, 6.1 10.6 10.6 1.0 0.0, 3.0 4.6 45.6 9.8 0.0, 20.9

Abbreviations: CI, confidence interval; NC, North Carolina; SMR, standardized mortality ratio.

a Number of suicide deaths per 100,000 person-years in the NC general population.

b Standardized number of suicide deaths per 100,000 person-years among persons formerly incarcerated in NC.

c 3-year rate, adjusted for sex, age, and year of release.

d 3-year rate, adjusted for age and year of release.

e 1-year rate, adjusted for sex and year of release.

SMRs comparing suicide mortality in the formerly incarcerated with general-population suicide mortality were variable by firearm involvement. Formerly incarcerated people had approximately 1.2 times the firearm-involved suicide mortality of the general population for the first 2 years after release, and then 1.4 (95% CI: 1.2, 1.6) times in the 3-year postrelease period. In contrast, formerly incarcerated people had 2.8–3.0 times the rate of non–firearm-involved suicide mortality of the general population for 3 years postrelease (Table 2). Non–firearm-involved suicide mortality included hanging, strangulation, suffocation, intentional overdose, and jumping (Web Table 1). Formerly incarcerated White and Hispanic/Latino people had more elevated suicide mortality compared with their general-population peers (SMR = 3.0 (95% CI: 2.7, 3.3) and SMR = 3.1 (95% CI: 1.3, 4.8), respectively) than did Black and Native American people (SMR = 1.3 (95% CI: 1.1, 1.6) and SMR = 1.2 (95% CI: 0.1, 2.2), respectively). White people in both the general population (21.4 per 100,000 PY) and the formerly incarcerated population (64.8 per 100,000 PY) had a higher suicide mortality rate than any other racial/ethnic group except formerly incarcerated Asian people (84.5 per 100,000 PY). The high suicide mortality rate and high SMR (SMR = 9.0, 95% CI: 0.0, 21.4) in Asian people were based on sparse data (2 suicide deaths) and should be interpreted cautiously. In terms of excess deaths, calculated as the observed number of deaths minus the expected number of deaths over the 3-year follow-up period, White people had the highest number (228.4), followed by Black people (25.1), Hispanic/Latino people (8.1), Asian people (1.8), and Native American people (0.8) (Figure 1, Table 2). Misclassification of race/ethnicity was observed, which may have biased the general-population rate calculations (Web Appendix 2).

Figure 1.

Figure 1

Standardized 1-year postrelease suicide mortality rate (number of deaths per 100,000 person-years) among formerly incarcerated people, suicide mortality rate in the general population, and associated standardized mortality ratios (SMRs), by year, North Carolina, 2000–2018. Bars, 95% confidence intervals.

In both the general population and the formerly incarcerated population, males had higher adjusted suicide mortality rates than females. However, SMRs in formerly incarcerated females compared with general-population females were higher (SM = 3.2, 95% CI: 2.1, 4.4) than in males (SMR = 1.9, 95% CI: 1.7, 2.1). In fact, the overall standardized suicide mortality rate of formerly incarcerated females (25.8 per 100,000 PY) was greater than that of males in the general population (24.9 per 100,000 PY), and the non–firearm-involved standardized suicide mortality rate of formerly incarcerated females (18.8 per 100,000 PY) was nearly double that of males in the general population (9.5 per 100,000 PY) (Table 2).

In the general population, overall suicide mortality rates were lowest in the age group 18–21 years (11.4 per 100,000 PY) and highest in the age group 50–59 years (19.0 per 100,000 PY). However, in the formerly incarcerated population, overall suicide rates were highest in the age groups 18–21 years (35.5 per 100,00 PY) and 60–69 years (46.7 per 100,000 PY). As a result, suicide mortality among emerging adults released from incarceration was 3 times that of their general-population peers, while suicide mortality for adults aged 22–49 years was approximately 2 times that of their general-population peers. SMR and standardized mortality rate estimates for formerly incarcerated people over age 50 years were based on sparse data (Table 2).

Suicide mortality rates were more variable by calendar year for the formerly incarcerated population as compared with the general population, resulting in variable 1-year SMRs by year of release. There were notable elevations of suicide mortality rates in individuals released in 2006 (45.4 per 100,000 PY), 2008 (48.4 per 100,000 PY), and 2018 (58.2 per 100,000 PY). In these years, formerly incarcerated people were 2.7 (95% CI: 1.3, 4.1), 2.9 (95% CI: 1.5, 4.2), and 3.1 (95% CI: 1.5, 4.7) times as likely to die by suicide in the 1-year postrelease period compared with the general population. In 2002 and 2012, formerly incarcerated people had equal or lower 1-year suicide mortality rates compared with the general population, with SMRs of 1.0 (95% CI: 0.02, 2.0) and 0.7 (95% CI: 0.0, 1.4), respectively (Figure 2).

Figure 2.

Figure 2

Three-year postrelease adjusted suicide mortality rate (number of deaths per 100,000 person-years) among formerly incarcerated people, suicide mortality rate in the general population, and excess deaths, by race/ethnicity, North Carolina, 2000–2020.

DISCUSSION

Study findings

Our study, carried out over nearly 2 decades among 266,400 formerly incarcerated people in North Carolina, demonstrated that these individuals are at a substantially elevated risk of suicide mortality compared with the general population for several years following their release. This finding of sustained elevated risk differs from that of other causes of death, like opioid overdose and all-cause mortality, for which the highest risk of death is during the 2 weeks immediately after release but elevated risk declines with time (6, 21, 22). However, our overall finding of elevated suicide mortality rates in formerly incarcerated people is consistent with earlier studies (4–6).

There are numerous mechanisms through which this elevated risk may be created and sustained. First, incarcerated individuals are known to have a high prevalence of disability, including physical, intellectual, and developmental disabilities, as well as mental health conditions and substance use disorders, which may contribute both to their becoming incarcerated and to their elevated risk of suicide mortality after release (32–37). Second, the experience of incarceration and the prison environment itself can contribute to the development of new mental health problems or exacerbation of existing mental health problems through solitude, lack of privacy, lack of meaningful activity, violence victimization, and insecurity about the future after release (23, 34, 38). While incarcerated, individuals have fewer opportunities to attempt suicide due to lack of privacy and lack of access to lethal means. They may be involved in mental health treatment, educational programs, and spiritual activities that may partially mitigate their risk of suicide while incarcerated. However, upon release they face a persistent lack of social, economic, and health-care support, as well as burdens of stigma, employment and housing discrimination, and debt recovery, which may contribute to psychological distress and suicidality during and after reentry, especially for individuals with preexisting disabilities (19, 37, 39–41). Further research examining these and other predictors’ effects on postrelease suicide mortality is ongoing (26, 42).

Formerly incarcerated people had higher rates of nonfirearm suicide mortality than firearm suicide mortality, which is contrary to trends in the general population. This is explainable in part by demographic differences between the two populations. People who have been incarcerated are younger and less often White, both characteristics associated with lower firearm suicide rates in the general population. However, this finding may also be explained by federal and state prohibition of firearm sales to individuals who have been incarcerated and/or convicted of a felony (24, 25). This is particularly well illuminated in the elevated nonfirearm suicide mortality rates among White and male formerly incarcerated people, whose demographics would typically indicate elevated firearm suicide mortality instead. Because formerly incarcerated people more often rely on less lethal means, their rates of nonfatal suicide attempt may be even more elevated compared with the general population than the rates of fatal suicide attempt demonstrated in this study; this should be investigated in future research.

We observed effect modification by race/ethnicity that demonstrated White and Hispanic/Latino people had a larger elevation in the suicide mortality rate after incarceration compared with Black people. Notably, excess deaths among formerly incarcerated White people were 9 times those of formerly incarcerated Black people even though there were fewer White people than Black people in the total formerly incarcerated population. This finding may seem paradoxical, given the established relationships between Blackness, incarceration, and health. Black people have been disproportionately incarcerated (43) (Table 1) due to criminal legal policies and practices in the United States—including proactive policing tactics typically concentrated in historically redlined and currently disinvested areas (15, 44, 45)—that result in greater frequency of police contact compared with White people (46–49), as well as disparate impacts of mandatory minimum sentencing and sentence enhancement policies (50–54). Incarceration has been shown to have adverse mental health effects during incarceration and after release (19, 23, 29, 55–57). There may be several avenues through which formerly incarcerated Black people are paradoxically less affected by postrelease suicide mortality. First, they may experience better support and acceptance from family and other social networks which regard the US system of incarceration as a force of oppression rather than regarding their loved one as criminal (58, 59). Second, formerly incarcerated Black people may experience “disadvantage saturation,” meaning the negative health effects of structural racism have already occurred prior to incarceration and consequently there is less of a compounding effect of incarceration on postrelease health in comparison with other groups (38, 60, 61). In contrast, formerly incarcerated White people may perceive a greater loss of social and economic status after incarceration (61), resulting in greater elevation of their already relatively high suicide mortality rates.

Further effect modification was observed by sex, age, and year of release from incarceration. The SMR comparing the formerly incarcerated with the general population was higher among females, though their adjusted suicide mortality rate remained approximately half the rate of formerly incarcerated males, which is consistent with prior literature (62). The relatively more elevated SMR in females may be explained by the fact that gendered pathways to incarceration are such that females are more likely to have traumatic backgrounds (63–65) and histories of mental illness (32, 64) and substance-use disorders (66) while experiencing heavy burdens of intersectional stigma upon release (67). Released emerging adults aged 18–21 years also had particularly elevated suicide mortality rates compared with their general-population peers, which may be due to the interruption in social network and identity development that occurs during this period (68, 69). This demonstrates a critical need for age-specific interventions, as these deaths represent a considerable loss of potential life.

There were some limitations to our analysis. First, our study consisted only of individuals released from North Carolina state prisons, and therefore our findings may not be generalizable to other US prison systems or to individuals experiencing other forms of criminal legal involvement within North Carolina. However, our findings will be useful for the purposes of comparison and meta-analysis across systems, since other large studies of formerly incarcerated people are ongoing (26). Additionally, data on individuals who migrated from North Carolina and died of suicide in other locations were not captured, resulting in an underestimate of suicide mortality in formerly incarcerated people. Given that the socioeconomic impacts of incarceration and restrictions from postrelease community supervision programs prohibit migration, we expect this underestimation to have been minimal, particularly for the first few years after release. Misclassification of suicide deaths in death certificate records is likely to exist and to be racially patterned (29, 70). Further, we found that misclassification of race/ethnicity in either death certificates or DAC records was evident for formerly incarcerated people (Web Appendix 2). Finally, linkage of DAC records to death certificate records is likely to be imperfect, but not to an extent that would negate our findings.

Conclusions and implications

Formerly incarcerated people in the United States are a particularly vulnerable and disproportionately Black and disabled population whose health trajectories are driven by structural barriers to housing, employment, and health care that influence and are influenced by their criminal legal involvement. As such, efforts to reduce suicide mortality in formerly incarcerated people should be rooted in antiracism, antiableism, and acknowledgement of the disparate harms of incarceration in vulnerable individuals. One effort by the North Carolina DAC that acknowledges these disparate harms is therapeutic and rehabilitative diversion units, which provide short-term specialized care to incarcerated individuals with high mental health needs who would otherwise be assigned to restrictive housing (or solitary confinement). Therapeutic and rehabilitative diversion units have been found to reduce inpatient mental health admissions and self-injury rates during the program (71). However, longer-term results show that self-harm rates remain elevated among treated individuals after the program ends (72). Further, across state prison systems, suicide mortality rates among incarcerated people who remain in custody have been steadily rising despite increased awareness of the mental health needs of incarcerated people (73).

Our findings of sustained suicide mortality risk in formerly incarcerated people for years following release, coupled with the limited long-term efficacy of ongoing interventions, suggest a need for more sustained and consistent support while individuals are incarcerated and during release planning, reentry, and community reintegration. Importantly, intervention efforts must go beyond those that prisons can provide. Pertinent action steps include programs designed to nurture development and maintenance of family and other social support networks, reduce barriers to housing and employment, expand community-based transition and support programs, restore voting rights, improve access to affordable health care, and prevent incarceration itself (74, 75).

Supplementary Material

Web_Material_kwad214
web_material_kwad214.pdf (147.1KB, pdf)

ACKNOWLEDGMENTS

Author affiliations: Department of Epidemiology, Gillings School of Global Public Health, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States (Kate Vinita Fitch, Brian W. Pence, Bradley N. Gaynes, Monica E. Swilley-Martinez, Shabbar I. Ranapurwala); Injury Prevention Research Center, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States (Kate Vinita Fitch, Vanessa E. Miller, Monica E. Swilley-Martinez, Shabbar I. Ranapurwala); Department of Medicine, School of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States (David L. Rosen, Timothy S. Carey); Department of Psychiatry, School of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States (Bradley N. Gaynes); Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States (Andrew L. Kavee, Timothy S. Carey); and Injury and Violence Prevention Branch, Division of Public Health, North Carolina Department of Health and Human Services, Raleigh, North Carolina, United States (Scott K.Proescholdbell).

This study was funded by the National Institute of Mental Health, National Institutes of Health (grant R01MH124752; B.W.P. and S.I.R., Multiple Principal Investigators). In addition, S.I.R. and D.L.R. were supported by the UNC Injury Prevention Research Center (grant R49CE003092; Stephen Marshall, Principal Investigator), and S.I.R. was also supported by the Justice Community Opioid Innovation Network (grant U01DA050442; Rosemarie Martin, Lauren Brinkley-Rubinstein, and Damaris Rohsenow, Multiple Principal Investigators).

The data used in this study are not publicly available but can be obtained upon request from the entities noted under “Data Sources” in the Methods section of this article.

We thank the staff of the North Carolina Department of Adult Corrections for their assistance.

This work was presented at the 56th Annual Meeting of the Society for Epidemiologic Research, Portland, Oregon, June 13–16, 2023.

The views expressed in this article are solely those of the authors.

Conflict of interest: none declared.

REFERENCES

  • 1. Ehlman  DC, Yard  E, Stone  DM, et al.  Changes in suicide rates—United States, 2019 and 2020. MMWR Morb Mortal Wkly Rep.  2022;71(8):306–312 [DOI] [PubMed] [Google Scholar]
  • 2. North Carolina Department of Health and Human Services . NC Violent Death Reporting System. https://dashboards.ncdhhs.gov/t/DPH/views/NCVDRSDashboard/NC-VDRS  Dashboard?%3Aembed=y&%3AisGuestRedirectFrom  Vizportal=y. Published November 2021. Accessed July 26, 2022.
  • 3. Janca  E, Keen  C, Willoughby  M, et al.  Sex differences in suicide, suicidal ideation, and self-harm after release from incarceration: a systematic review and meta-analysis. Soc Psychiatry Psychiatr Epidemiol.  2022;58(3):355–371. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Rosen  DL, Schoenbach  VJ, Wohl  DA. All-cause and cause-specific mortality among men released from state prison, 1980–2005. Am J Public Health.  2008;98(12):2278–2284. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Lize  SE, Scheyett  AM, Morgan  CR, et al.  Violent death rates and risk for released prisoners in North Carolina. Violence Vict.  2015;30(6):1019–1036. [DOI] [PubMed] [Google Scholar]
  • 6. Binswanger  IA, Stern  MF, Deyo  RA, et al.  Release from prison—a high risk of death for former inmates. N Engl J Med.  2007;356(2):157–165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Pratt  D, Piper  M, Appleby  L, et al.  Suicide in recently released prisoners: a population-based cohort study. Lancet.  2006;368(9530):119–123. [DOI] [PubMed] [Google Scholar]
  • 8. Zlodre  J, Fazel  S. All-cause and external mortality in released prisoners: systematic review and meta-analysis. Am J Public Health.  2012;102(12):e67–e75. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Bukten  A, Stavseth  MR. Suicide in prison and after release: a 17-year national cohort study. Eur J Epidemiol.  2021;36(10):1075–1083. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Liu  YE, Lemos  EF, Gonçalves  CCM, et al.  All-cause and cause-specific mortality during and following incarceration in Brazil: a retrospective cohort study. PLoS Med.  2021;18(9):e1003789. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Morgan  ER, Rivara  FP, Ta  M, et al.  Incarceration and subsequent risk of suicide: a statewide cohort study. Suicide Life Threat Behav.  2022;52(3):467–477. [DOI] [PubMed] [Google Scholar]
  • 12. North Carolina Department of Adult Corrections . DAC research and planning. Automated System Query (A. S. Q. DOC 3.0b). https://webapps.doc.state.nc.us/apps/asqExt/ASQ. Accessed June 25, 2023.
  • 13. North Carolina Department of Public Safety . Prison reform in North Carolina. https://www.ncdps.gov/about-dps/current-initiatives/prison-reform-north-carolina. Accessed June 26, 2023.
  • 14. North Carolina General Assembly . Article 84A. Post-Release Supervision. (General Statutes of North Carolina §15A-1368). https://www.ncleg.net/EnactedLegislation/Statutes/PDF/ByArticle/Chapter_15a/Article_84A.pdf. Accessed September 11, 2023.
  • 15. Alexander  M. The New Jim Crow: Mass Incarceration in the Age of Colorblindness. New York, NY: The New Press; 2010. [Google Scholar]
  • 16. Wacquant  L. Class, race & hyperincarceration in revanchist America. Daedalus.  2010;139(3):74–90. [Google Scholar]
  • 17. Wakefield  S, Uggen  C. Incarceration and stratification. Annu Rev Sociol.  2010;36:387–406. [Google Scholar]
  • 18. Bailey  ZD, Feldman  JM, Bassett  MT. How structural racism works—racist policies as a root cause of U.S. racial health inequities. N Engl J Med.  2021;384(8):768–773. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Brinkley-Rubinstein  L. Incarceration as a catalyst for worsening health. Health Justice.  2013;1(1):1–17. [Google Scholar]
  • 20. Brinkley-Rubinstein  L, Sivaraman  J, Rosen  DL, et al.  Association of restrictive housing during incarceration with mortality after release. JAMA Netw Open.  2019;2(10):e1912516. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Ranapurwala  SI, Shanahan  ME, Alexandridis  AA, et al.  Opioid overdose mortality among former North Carolina inmates: 2000–2015. Am J Public Health.  2018;108(9):1207–1213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Ranapurwala  SI, Figgatt  MC, Remch  M, et al.  Opioid overdose deaths among formerly incarcerated persons and the general population: North Carolina, 2000–2018. Am J Public Health.  2022;112(2):300–303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Schnittker  J, Massoglia  M, Uggen  C. Out and down. J Health Soc Behav.  2012;53(4):448–464. [DOI] [PubMed] [Google Scholar]
  • 24. North Carolina General Assembly . Article 54A. The Felony Firearms Act. (General Statutes of North Carolina §14-415). https://www.ncleg.gov/EnactedLegislation/Statutes/PDF/BySection/Chapter_14/GS_14-415.1.pdf. Accessed September 11, 2023.
  • 25. Congress of the United States . 18 U.S. Code §922—Unlawful Acts. (U.S. Code, Title 18, Part I, Chapter 44, Section 922). https://www.law.cornell.edu/uscode/text/18/922. Accessed November 17, 2022.
  • 26. Borschmann  R, Tibble  H, Spittal  MJ, et al.  The Mortality After Release from Incarceration Consortium (MARIC): protocol for a multi-national, individual participant data meta-analysis. Int J Popul Data Sci.  2020;5(1):1145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. National Center for Health Statistics . U.S. Census populations with bridged race categories. (Vintage 2020 postcensal estimates of the resident population of the United States (April 1, 2010, July 1, 2010–July 1, 2020), by year, county, single-year of age (0, 1, 2, ..., 85 years and over), bridged race, Hispanic origin, and sex. Prepared under a collaborative arrangement with the U.S. Census Bureau as of September 22, 2021, following release by the U.S. Census Bureau of the unbridged vintage 2020 postcensal estimates by 5-year age group on June 17, 2021). http://www.cdc.gov/nchs/nvss/bridged_race.htm. Published 2021. Accessed October 3, 2023.
  • 28. National Center for Health Statistics . Bridged-race population estimates—data files and documentation. (Intercensal estimates of the resident population of the United States for July 1, 2000–July 1, 2009, by year, county, single-year of age (0, 1, 2, ..., 85 years and over), bridged race, Hispanic origin, and sex). https://www.cdc.gov/nchs/nvss/bridged_race/data_documentation.htm. Published 2012. Accessed October 3, 2023.
  • 29. Rockett  IRH, Wang  S, Stack  S, et al.  Race/ethnicity and potential suicide misclassification: window on a minority suicide paradox?  BMC Psychiatry.  2010;10(1):1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Arias  E, Heron  M, Hakes  J. The validity of race and Hispanic-origin reporting on death certificates in the United States: an update. Vital Health Stat 2.  2016;(172):1–21. [PubMed] [Google Scholar]
  • 31. Westreich  D, Cole  SR. Invited commentary: positivity in practice. Am J Epidemiol.  2010;171(6):674–677. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Fazel  S, Seewald  K. Severe mental illness in 33 588 prisoners worldwide: systematic review and meta-regression analysis. Br J Psychiatry.  2012;200(5):364–373. [DOI] [PubMed] [Google Scholar]
  • 33. Prins  SJ. Prevalence of mental illnesses in U.S. state prisons: a systematic review. Psychiatr Serv.  2014;65(7):862–872. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Durcan  G, Zwemstra  JC. Mental health in prison. In: Enggist  S, Møller  L, Galea  G, et al., eds. Prisons and Health. Geneva, Switzerland: World Health Organization; 2014:87–95. https://apps.who.int/iris/bitstream/handle/10665/128603/9789289050593-eng.pdf?sequence=3&isAllowed=y. Accessed August 23, 2022. [Google Scholar]
  • 35. Baillargeon  J, Binswanger  IA, Penn  J, et al.  Psychiatric disorders and repeat incarcerations: the revolving prison door. Am J Psychiatry.  2009;166(1):103–109. [DOI] [PubMed] [Google Scholar]
  • 36. Bixby  L, Bevan  S, Boen  C. The links between disability, incarceration, and social exclusion. Health Aff.  2022;41(10):1460–1469. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Leotti  SM, Slayter  E. Criminal legal systems and the disability community: an overview. Soc Sci.  2022;11(6):255. [Google Scholar]
  • 38. Sugie  NF, Turney  K. Beyond incarceration: criminal justice contact and mental health. Am Sociol Rev.  2017;82(4):719–743. [Google Scholar]
  • 39. Link  NW. Criminal justice debt during the prisoner reintegration process: who has it and how much?  Crim Justice Behav.  2018;46(1):154–172. [Google Scholar]
  • 40. Semenza  DC, Link  NW. How does reentry get under the skin? Cumulative reintegration barriers and health in a sample of recently incarcerated men. Soc Sci Med.  2019;243:112618. [DOI] [PubMed] [Google Scholar]
  • 41. Addison  HA, Richmond  TS, Lewis  LM, et al.  Mental health outcomes in formerly incarcerated black men: a systematic mixed studies review. J Adv Nurs.  2022;78(7):1851–1869. [DOI] [PubMed] [Google Scholar]
  • 42. Ranapurwala  SI, Miller  VE, Carey  TS, et al.  Innovations in suicide prevention research (INSPIRE): a protocol for a population-based case–control study. Inj Prev.  2022;28(5):483–490. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Nellis  A. The color of justice: racial and ethnic disparity in state prisons. https://www.sentencingproject.org/reports/the-color-of-justice-racial-and-ethnic-disparity-in-state-prisons-the-sentencing-project/. Published October 13, 2021. Accessed November 28, 2022.
  • 44. Mitchell  J, Chihaya  GK. Tract level associations between historical residential redlining and contemporary fatal encounters with police. Soc Sci Med.  2022;302:114989. [DOI] [PubMed] [Google Scholar]
  • 45. Appel  I, Nickerson  J. Pockets of poverty: the long-term effects of redlining. 10.2139/SSRN.2852856. Published October 17, 2016. Accessed June 27, 2023. [DOI]
  • 46. Pierson  E, Simoiu  C, Overgoor  J, et al.  A large-scale analysis of racial disparities in police stops across the United States. Nat Hum Behav.  2020;4(7):736–745. [DOI] [PubMed] [Google Scholar]
  • 47. Gelman  A, Fagan  J, Kiss  A. An analysis of the New York City Police Department’s ‘stop-and-frisk’ policy in the context of claims of racial bias. J Am Stat Assoc.  2007;102(479):813–823. [Google Scholar]
  • 48. Kubrin  CE, Messner  SF, Deane  G, et al.  Proactive policing and robbery rates across U.S. cities. Criminology.  2010;48(1):57–97. [Google Scholar]
  • 49. Geller  A, Fagan  J. Pot as pretext: marijuana, race, and the new disorder in New York City street policing. J Empir Leg Stud. 2010;7(4):591–633. [Google Scholar]
  • 50. Starr  SB, Rehavi  MM. Mandatory sentencing and racial disparity: assessing the role of prosecutors and the effects of Booker. Yale Law J.  2013;123(1):1–77. [Google Scholar]
  • 51. Marit Rehavi  M, Starr  SB. Racial disparity in federal criminal sentences. J Polit Econ.  2014;122(6):1320–1354. [Google Scholar]
  • 52. Fischman  JB, Schanzenbach  MM. Racial disparities under the federal sentencing guidelines: the role of judicial discretion and mandatory minimums. J Empir Leg Stud.  2012;9(4):729–764. [Google Scholar]
  • 53. Mustard  DB. Racial, ethnic, and gender disparities in sentencing: evidence from the U.S. federal courts. J Law Econ.  2001;44(1):285–314. [Google Scholar]
  • 54. Schlesinger  T. The failure of race neutral policies: how mandatory terms and sentencing enhancements contribute to mass racialized incarceration. Crime Delinq.  2011;57(1):56–81. [Google Scholar]
  • 55. Haney  C. Mental health issues in long-term solitary and “supermax” confinement. Crime Delinq.  2003;49(1):124–156. [Google Scholar]
  • 56. James  DJ, Glaze  LE. Mental Health Problems of Prison and Jail Inmates. (Bureau of Justice Statistics Special Report). Washington, DC: US Department of Justice; 2006. https://bjs.ojp.gov/content/pub/pdf/mhppji.pdf. Accessed August 24, 2022. [Google Scholar]
  • 57. Wildeman  C, Wang  EA. Mass incarceration, public health, and widening inequality in the USA. Lancet.  2017;389(10077):1464–1474. [DOI] [PubMed] [Google Scholar]
  • 58. Unnever  JD. Two worlds far apart: black-white differences in beliefs about why African-American men are disproportionately imprisoned. Criminology.  2008;46(2):511–538. [Google Scholar]
  • 59. Elliott  S, Reid  M. Low-income black mothers parenting adolescents in the mass incarceration era: the long reach of criminalization. Am Sociol Rev.  2019;84(2):197–219. [Google Scholar]
  • 60. Hannon  L. Poverty, delinquency, and educational attainment: cumulative disadvantage or disadvantage saturation?  Sociol Inq.  2003;73(4):575–594. [Google Scholar]
  • 61. Massoglia  M, Firebaugh  G, Warner  C. Racial variation in the effect of incarceration on neighborhood attainment. Am Sociol Rev.  2013;78(1):142–165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62. Smith  M, Udo  T. The relationship between age at incarceration and lifetime suicide attempt among a nationally representative sample of U.S. adults. Community Ment Health J.  2022;58(7):1403–1415. [DOI] [PubMed] [Google Scholar]
  • 63. Boehm  J, Oldfield  D, Martin  F, et al.  Gendered pathways: a quantitative investigation of women probationers’ paths to incarceration. Crim Justice Behav.  2009;36(6):541–566. [Google Scholar]
  • 64. Karlsson  ME, Zielinski  MJ. Sexual victimization and mental illness prevalence rates among incarcerated women: a literature review. Trauma Violence Abuse.  2020;21(2):326–349. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65. Baranyi  G, Cassidy  M, Fazel  S, et al.  Prevalence of posttraumatic stress disorder in prisoners. Epidemiol Rev.  2018;40(1):134–145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66. Fazel  S, Yoon  IA, Hayes  AJ. Substance use disorders in prisoners: an updated systematic review and meta-regression analysis in recently incarcerated men and women. Addiction.  2017;112(10):1725–1739. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67. Gunn  AJ, Sacks  TK, Jemal  A. “That’s not me anymore”: resistance strategies for managing intersectional stigmas for women with substance use and incarceration histories. Qual Soc Work.  2016;17(4):490–508. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68. Pettus-Davis  C, Doherty  EE, Veeh  C, et al.  Deterioration of postincarceration social support for emerging adults. Crim Justice Behav.  2017;44(10):1317–1339. [Google Scholar]
  • 69. Arditti  JA, Parkman  T. Young men’s reentry after incarceration: a developmental paradox. Fam Relat.  2011;60(2):205–220. [Google Scholar]
  • 70. Ali  B, Rockett  IRH, Miller  TR, et al.  Racial/ethnic differences in preceding circumstances of suicide and potential suicide misclassification among US adolescents. J Racial Ethn Health Disparities.  2022;9(1):296–304. [DOI] [PubMed] [Google Scholar]
  • 71. Remch  M, Mautz  C, Burke  EG, et al.  Impact of a prison therapeutic diversion unit on mental and behavioral health outcomes. Am J Prev Med.  2021;61(5):619–627. [DOI] [PubMed] [Google Scholar]
  • 72. Remch  M, Mautz  C, Austin  AE, et al.  Sustained impacts of North Carolina prison therapeutic diversion units on behavioral outcomes, mental health, self-injury, and restrictive housing readmission. Prev Med.  2022;164:107318. [DOI] [PubMed] [Google Scholar]
  • 73. Carson  EA. Mortality in state and federal prisons, 2001–2018—statistical tables. Washington, DC: US Department of Justice; 2021. https://bjs.ojp.gov/content/pub/pdf/msfp0118st.pdf. Accessed January 22, 2023. [Google Scholar]
  • 74. American Public Health Association . Advancing Public Health Interventions to Address the Harms of the Carceral System. https://www.apha.org/Policies-and-Advocacy/Public-Health-Policy-Statements/Policy-Database/2022/  01/07/Advancing-Public-Health-Interventions-to-Address-the-Harms-of-the-Carceral-System. (Policy no. 202117). Published October 26, 2021. Accessed January 10, 2023.
  • 75. Wallace  D, Fahmy  C, Cotton  L, et al.  Examining the role of familial support during prison and after release on post-incarceration mental health. Int J Offender Ther Comp Criminol.  2016;60(1):3–20. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Web_Material_kwad214
web_material_kwad214.pdf (147.1KB, pdf)

Articles from American Journal of Epidemiology are provided here courtesy of Oxford University Press

RESOURCES