JAMAIOpen,.

Network

Original Investigation | Public Health

Incarceration of Youths in an Adult Correctional Facility and Risk of Premature Death
Ian A. Silver, PhD; Daniel C. Semenza, PhD; Joseph L. Nedelec, PhD

Abstract
IMPORTANCE Youths incarcerated in adult correctional facilities are exposed to a variety of adverse
circumstances that could diminish psychological and physical health, potentially leading to early
mortality.

Key Points
Question Is incarceration of youths in
adult correctional facilities associated
with an increased risk of mortality
through 39 years of age?

OBJECTIVE To evaluate whether being incarcerated in an adult correctional facility as a youth was

Findings In this cohort study of 8951

associated with mortality between 18 and 39 years of age.

youths, incarceration in an adult
correctional facility before the age of 18

DESIGN, SETTING, AND PARTICIPANTS This cohort study relied on longitudinal data collected

years was associated with a 33%

from 1997 to 2019 as part of the National Longitudinal Survey of Youth–1997, a nationally

increase in the risk of mortality between

representative sample of 8984 individuals born in the United States between January 1, 1980, and

18 and 39 years of age.

December 1, 1984. The data analyzed for the current study were derived from annual interviews
between 1997 and 2011 and interviews every other year from 2013 to 2019 (19 interviews in total).
Participants were limited to respondents aged 17 years or younger during the 1997 interview and
alive during their 18th birthday (8951 individuals; >99% of the original sample). Statistical analysis
was performed from November 2022 to May 2023.

Meaning This study suggests that
incarceration in an adult correctional
facility as a youth was associated with
early mortality, potentially through
diminished psychological and
physical health.

INTERVENTION Incarceration in an adult correctional facility before the age of 18 years compared
with being arrested before the age of 18 years or never arrested or incarcerated before the age of
18 years.
MAIN OUTCOMES AND MEASURES The main outcome for the study was age at mortality between
18 and 39 years of age.

+ Invited Commentary
+ Supplemental content
Author affiliations and article information are
listed at the end of this article.

RESULTS The sample of 8951 individuals included 4582 male participants (51%), 61 American Indian
or Alaska Native participants (1%), 157 Asian participants (2%), 2438 Black participants (27%), 1895
Hispanic participants (21%), 1065 participants of other race (12%), and 5233 White participants
(59%). A total of 225 participants (3%) died during the study period, with a mean (SD) age at death
of 27.7 (5.9) years. Incarceration in an adult correctional facility before the age of 18 years was
associated with an increased risk of earlier mortality between 18 and 39 years of age compared with
individuals who were never arrested or incarcerated before the age of 18 years (time ratio, 0.67; 95%
CI, 0.47-0.95). Being arrested before the age of 18 years was associated with an increased risk of
earlier mortality between 18 and 39 years of age when compared with individuals who were never
arrested or incarcerated before the age of 18 years (time ratio, 0.82; 95% CI, 0.73-0.93).
CONCLUSIONS AND RELEVANCE In this cohort study of 8951 youths, the survival model suggested
that being incarcerated in an adult correctional facility may be associated with an increased risk of
early mortality between 18 and 39 years of age.
JAMA Network Open. 2023;6(7):e2321805.
Last corrected on July 24, 2023. doi:10.1001/jamanetworkopen.2023.21805

Open Access. This is an open access article distributed under the terms of the CC-BY License.
JAMA Network Open. 2023;6(7):e2321805. doi:10.1001/jamanetworkopen.2023.21805 (Reprinted)

Downloaded From: https://jamanetwork.com/ on 08/02/2023

July 5, 2023

1/10

Downloaded From: https://jamanetwork.com/ on 08/02/2023

JAMA Network Open. 2023;6(7):e2321805. doi:10.1001/jamanetworkopen.2023.21805 (Reprinted)

The approval for the data collection was obtained by the institutional review boards at The Ohio

approval as the study relied on publicly available deidentified data that can be retrieved by anyone.

reporting on a secondary publicly available cohort study. We did not seek institutional review board

(STROBE) reporting guideline for reporting by satisfying all the applicable items (1-21) in terms of

The current study follows the Strengthening the Reporting of Observational Studies in Epidemiology

Analytical Sample

Methods

via arrests, associated risk factors, and demographic differences.

with greater risk for early mortality after accounting for broader exposure to the criminal legal system

We hypothesize that juvenile incarceration in an adult correctional facility will be associated

with early mortality after release for those who have been incarcerated as a youth in an adult facility.

victimization, and higher engagement in risky behaviors, such as substance abuse, may be associated

and developmental processes. Poorer mental and physical health, greater risk for violence and

youths as a result of more harmful exposures in prison and disruption to key social, psychological,

developmental periods.23 Incarceration among adults is likely to have damaging iatrogenic effects for

and violence, which can lead to further isolation and loss of socialization during critical

facilities often struggle to adjust to prison life and display heightened rates of disciplinary misconduct

environments, exposing them to greater risk for sexual and physical assault.18,21,22 Youths in adult

Youths incarcerated in adult facilities are more likely to engage with antisocial adult peers in prison

Young people housed in adult correctional facilities report being more afraid for their safety.14

were associated with early mortality among a sample of youths in Indiana.12

to the present study, researchers have observed that detention in jail and transfer to adult court

more distress and higher rates of psychiatric symptoms than those in juvenile centers.18 Most salient

depression compared with those in juvenile facilities.18-20 In Texas, youths in adult facilities reported

incarcerated in adult facilities report substantially greater rates of posttraumatic stress disorder and

documents serious extralegal consequences for youths incarcerated in adult facilities. Youths

of formally processed juvenile delinquency cases are transferred to an adult criminal court,17 research

place less emphasis on treatment, counseling, and education.16 Even though only approximately 1%

different experiences. Adult facilities are often much larger, have higher resident to staff ratios, and

in adult prison facilities.14,15 Incarceration in juvenile vs adult correctional facilities represents vastly

In most US states, youths can be transferred and sentenced in adult court, resulting in detention

early death.13

reproductive health, risk-taking behaviors, and mental well-being that heighten the likelihood of

severe.12 Youths who are incarcerated experience health challenges related to dental care, sexual and

nonincarcerated youths as involvement in the criminal legal system becomes more protracted and

violence and homicide victimization.6 Incarcerated youths face greater risk for early death than

nonincarcerated Medicaid-enrolled youths of the same age, associated largely with exposure to

a history of incarceration have an all-cause mortality rate roughly 5.9 times higher than observed in

mortality specifically among youths, legally defined as children younger than 18 years.11 Youths with

A smaller body of research documents the association of incarceration exposure with early

infectious disease.9,10

victimization and homicide,6 substance use and overdose,7,8 and higher incidence of chronic and

have been shown to link incarceration to early mortality, including greater risk for violent

or roughly 13% of the average US life expectancy at the age of 45 years.5 Numerous mechanisms

Mortality excesses associated with incarceration translate to losses of life expectancy of 4 to 5 years,

association such that more time served in prison corresponds to greater reductions in life span.4

July 5, 2023

2/10

Incarceration of Youths in an Adult Correctional Facility and Risk of Premature Death

Incarceration exposure is associated with early mortality.1-3 Research documents a dose-response

Introduction

JAMA Network Open | Public Health

c..G

Downloaded From: https://jamanetwork.com/ on 08/02/2023

JAMA Network Open. 2023;6(7):e2321805. doi:10.1001/jamanetworkopen.2023.21805 (Reprinted)

in the survival model.27,28 These covariates include items measuring the self-rated general health of

known to be associated with contact with the criminal legal system and early mortality were included

formed, directed acyclic graph (eAppendix 1 and eFigure in Supplement 1).1,10,12,13 Four covariates

confounders—and exclude mediators—following the development of a theoretically and empirically

Study variables were selected to limit the potential bias associated with colliders and

Control Variables: General Risk Factors and Demographic Characteristics

facility (jail or prison) and the number of months spent in an adult correctional facility.26

the current study capture data on whether an adolescent was incarcerated in an adult correctional

the monthly arrays for incarceration created by the NLSY97 research team. The measures created for

intoxication, pretrial detention, and confinement in juvenile detention centers were excluded from

offenses (eg, truancy), holds for DUI (driving under the influence of alcohol or drugs), holds for public

respondent experienced before the age of 18 years. Periods of confinement associated with status

created to measure the number of arrests and the number of months incarcerated that each

arrested or incarcerated before the age of 18 years (0 = no; 1 = yes). Two continuous variables were

the age of the respondent each year, 2 variables were created to measure whether a respondent was

or more days in the correctional facility during that month. With the use of the monthly arrays and

NLSY97 research team considered an individual incarcerated during a specified month if they spent 1

any point during a specified month from 1992 to 2019. Due to concerns associated with privacy, the

identifying the number of arrests a respondent experienced and if a respondent was incarcerated at

incarceration period since their last interview. The NLSY97 research team developed monthly arrays

The respondents were requested to report the date of each arrest and the start and end date for each

Key Independent Variables: Arrest and Incarceration in Adult Correctional Facilities

recorded in the age at death measure for individuals who did not die between 18 and 39 years of age.

Consistent with the implementation of a survival analysis, the age at the final interview was also

and confirm death records for all respondents using official data sources from state agencies.

the year they died, as reported by family members or friends. The NLSY97 research team did request

the US Social Security Office.25 Second, a respondent’s age at death was recorded as their age during

years of age. The mortality rate observed in the NLSY97 is consistent with the estimates produced by

(0 = no; 1 = yes). Of the 8951 individuals included in the sample, 225 (3%) died between 18 and 39

First, a variable for death was created to indicate if a respondent died between 18 and 39 years of age

nonresponse tracking for the NLSY97. With the use of this information, 2 variables were created.

Respondents who did not complete an interview because they died were identified as part of the

Dependent Variables: Death and Age at Death

Measures

any respondents.

sample). Because the study uses publicly available deidentified data, consent was not obtained from

during the 1997 interview and alive during their 18th birthday (8951 individuals; >99% of the original

to the respondent. The analytical sample included respondents who were 17 years of age or younger

the interview, the interview was completed in person in a visitation room or via telephone at no cost

and from 34 to 39 years during the 2019 interview. If a respondent was incarcerated at the time of

been completed, with the respondents ranging in age from 12 to 18 years during the 1997 interview

from 1997 to 2011 and in interviews every 2 years from 2013 to 2019. In total, 19 interviews have

1, 1980, and December 1, 1984.24 The 8984 respondents agreed to participate in yearly interviews

NLSY97, a nationally representative sample of individuals born in the United States between January

Labor Statistics. The data for the current study are derived from the publicly available version of the

National Longitudinal Study of Youth–1997 (NLSY97) is provided on the website of the US Bureau of

research team, with which we have no affiliation. All data documentation associated with the

July 5, 2023

3/10

Incarceration of Youths in an Adult Correctional Facility and Risk of Premature Death

State University and National Opinion Research Center at the University of Chicago by the NLSY97

JAMA Network Open | Public Health

c..G

Downloaded From: https://jamanetwork.com/ on 08/02/2023

JAMA Network Open. 2023;6(7):e2321805. doi:10.1001/jamanetworkopen.2023.21805 (Reprinted)

incarcerated for up to 3.5 years in an adult facility.

youths in an adult correctional facility. Some youths experienced as many as 17 arrests and were

sample were arrested before the age of 18 years, while 109 respondents (1%) were incarcerated as

mean (SD) age at death of 27.7 (5.9) years. A total of 1597 of the respondents (18%) in the analytical

participants (59%) (Table 1). A total of 225 respondents (3%) died during the study period, with a

Hispanic participants (ethnicity; 21%), 1065 participants of other race (12%), and 5233 White

or Alaska Native participants (1%), 157 Asian participants (2%), 2438 Black participants (27%), 1895

The analytical sample of 8951 individuals included 4582 male participants (51%), 61 American Indian

Results

analyses is provided in eAppendix 4 in Supplement 1.

4.2.3.36 To maintain open science, the R script used to clean the data and estimate all of the statistical

P < .05. All analyses were estimated using the survival35 and the SurvMetrics packages in R, version

results. All P values were from 2-sided tests and results were deemed statistically significant at

cumulative probability of death was calculated and plotted to permit a visual evaluation of the

death. The TR is calculated as the exponentiated value of the slope coefficient (exp [b]). The

than 1 indicate an increased odds of survival and values lower than 1 indicate an increased risk of

time ratio (TR) represents the risk of an event occurring across all time periods, where values higher

proportional hazard assumption of a Cox proportional hazards regression survival model.32-34 The

association between the independent variables and the dependent variable did not satisfy the

compared. A parametric survival model was preferred for the current analysis because the

Weibull, gaussian, logistic, and loglogistic—and the resulting Akaike information criteria were

intercept-only model was estimated with each distributional specification—lognormal, exponential,

was estimated using a lognormal distribution given the distributional properties of age at death. An

independent and control variables using a parametric survival model. The parametric survival model

descriptive statistics were produced. Third, death status and the age at death were regressed on the

arrests, the number of months incarcerated, or the death status of the respondents. Second,

version 4.2.3 [R Project for Statistical Computing]).31 No information was missing on the number of

data sets with 10 iterations each (pulled using the complete function from the mice package in R,

eTable 2 in Supplement 1).31 The imputed values for the covariates represent the first of 5 imputed

constructs) was implemented to impute missing values on the covariates (eAppendix 2, eTable 1, and

random forest method (continuous constructs) and the logistic regression method (dichotomous

was implemented. First, a missing data analysis was conducted, and multiple imputation using the

Statistical analysis was performed from November 2022 to May 2023. A 4-part analytical strategy

Statistical Analysis

to adjust for possible differences in the sampling procedures.24

respondent participated (0 = initial sample; 1 = supplemental sample) was also included in the model

Hispanic]). A dichotomous indicator (supplemental sample) for the NLSY97 sample in which the

individuals of multiple races and individuals who did not identify a race after initially identifying as

(American Indian or Alaska Native, Asian, Black, White; reference was other race [including

male sex (reference = female sex), ethnicity (Hispanic; reference = non-Hispanic), and race

In addition to these risk factors, demographic characteristics were adjusted for the inclusion of

18 years.

housing) before the age of 18 years, and the respondent’s household net worth before the age of

experienced homelessness, lived in a place without water or electricity, or resided in emergency

respondent turned 16 years of age, whether the respondent experienced childhood adversity (eg,

measured in 1997),29,30 whether 1 or both of the respondent’s parents were incarcerated before the

July 5, 2023

4/10

Incarceration of Youths in an Adult Correctional Facility and Risk of Premature Death

the respondent before the age of 18 years (0 = poor; 1 = fair; 2 = good; 3 = very good; 4 = excellent;

JAMA Network Open | Public Health

c..G

623 (7)
489 (5)
NA

Childhood adversity
Household net worth before the age of 18 y, $

4369 (49)

Female

7056 (79)

Non-Hispanic

2438 (27)
1065 (12)
5233 (58)

Black
Othera
White

0.25 (0.43)

0.58 (0.49)

0.12 (0.32)

0.27 (0.45)

0.02 (0.13)

0.006 (0.08)

0.79 (0.41)

0.21 (0.41)

0.49 (0.50)

0.51 (0.50)

92 288 (139 363)

0.05 (0.23)

0.07 (0.25)

3.06 (0.91)

0.38 (1.19)

0.17 (0.37)

0.08 (1.12)

0.01 (0.11)

27.7 (5.9)

0.03 (0.16)

Mean (SD)

Downloaded From: https://jamanetwork.com/ on 08/02/2023

JAMA Network Open. 2023;6(7):e2321805. doi:10.1001/jamanetworkopen.2023.21805 (Reprinted)

2230 (25)

157 (2)

Asian

Supplemental sample

61 (1)

American Indian or Alaska Native

Race

1895 (21)

Hispanic

Ethnicity

4582 (51)

Male

Sex

Demographics

NA

Parental incarceration before age 16 y

NA

1597 (18)

NA

109 (1)

NA

225 (3)

No. (%)
(N = 8951)

Self-rated health before age 18 y

General risk factors

No. of arrests

Arrested

No. of months incarcerated in adult facility

Incarcerated in adult facility

Key independent variables (before age 18 y)

Age at death, y

Death

Key dependent variable (aged 18-39 y)

Characteristic

Table 1. Descriptive Statistics for the Analytical Sample (N = 8951)

0 to 1

0 to 1

0 to 1

0 to 1

0 to 1

0 to 1

0 to 1

0 to 1

0 to 1

0 to 1

−935 251 to
600 000

0 to 1

0 to 1

0 to 4

0 to 17

0 to 1

0 to 42

0 to 1

18 to 39

0 to 1

Range

accounting for time served in a juvenile facility. The association between incarceration in an adult

(eAppendix 3, eTable 3, and eTable 4 in Supplement 1) assessed whether the findings held after

associated with an increased risk of death between 18 and 39 years of age. Supplemental analyses

legal system as a youth—both arrest and, especially, incarceration in adult correctional facilities—was

estimated to have died by the age of 39 years. Overall, the findings suggest that contact with the

years of age and just over 2% of youths without legal system contact before the age of 18 years were

estimated to die by the age of 39 years. In comparison, just over 5% of youths arrested before 18

shown in the Figure, approximately 8% of youths incarcerated in adult correctional facilities were

facilities, arrested youths, and respondents without legal system contact before 18 years of age. As

We plotted the cumulative probability of death for youths incarcerated in adult correctional

associated with a higher risk of early death (TR, 0.84; 95% CI, 0.77-0.91).

associated with a higher likelihood of survival (TR, 1.10; 95% CI, 1.05-1.15), while being male was

with the risk of death prior to 39 years of age. Better general health before 18 years of age was

(TR, 1.02; 95% CI, 0.99-1.06) nor the number of arrests (TR, 1.00; 95% CI, 0.97-1.04) were associated

compared with not being arrested. Neither the number of months of incarceration in an adult facility

associated with an increase in the risk of death until the age of 39 years (TR, 0.82; 95% CI, 0.73-0.93)

compared with nonincarcerated youths. In addition, being arrested before the age of 18 years was

associated with an approximate 33% increase in the risk of death (TR, 0.67; 95% CI, 0.47-0.95)

death before the age of 39 years. Incarceration of youths in an adult correctional facility was

a

July 5, 2023

5/10

Includes individuals of multiple races and individuals
who did not identify a race after initially identifying
as Hispanic.

Abbreviation: NA, not applicable.

Incarceration of Youths in an Adult Correctional Facility and Risk of Premature Death

Table 2 provides the results of the parametric survival model that assessed the likelihood of

JAMA Network Open | Public Health

c..G

1.00 (0.97 to 1.04)

−0.20 (0.06) [−0.32 to −0.07]
0.00 (0.02) [−0.03 to 0.04]

Arrestedb

−0.17 (0.07) [−0.31 to −0.04]
0.00 (0.09) [−0.17 to 0.17]
0.00 (0.00) [0.00 to 0.00]

Parental incarceration before 16 yb
Childhood adversity
Household net worth before 18 yb

0.27 (0.28) [−0.28 to 0.81]
−0.16 (0.10) [−0.36 to 0.04]
0.08 (0.07) [−0.06 to 0.22]
−0.07 (0.10) [−0.25 to 0.12]
0.05 (0.06) [−0.06 to 0.16]

Asian (reference: other)
Black (reference: other)
Hispanic (reference: non-Hispanic)
White (reference: other)
Supplemental sample

0.723

Scale

NA

18

21

24

27

Age, y

30

33

36

No CLS contact

Arrested <18 y

Incarcerated <18 y

Downloaded From: https://jamanetwork.com/ on 08/02/2023

JAMA Network Open. 2023;6(7):e2321805. doi:10.1001/jamanetworkopen.2023.21805 (Reprinted)

0

0.02

0.04

0.06

0.08

0.10

0.12

39

1.05 (0.94 to 1.17)

0.94 (0.78 to 1.13)

1.08 (0.94 to 1.25)

0.85 (0.70 to 1.04)

1.31 (0.76 to 2.26)

1.02 (0.56 to 1.84)

0.84 (0.77 to 0.91)

1.00 (1.00 to 1.00)

1.00 (0.84 to 1.18)

0.84 (0.74 to 0.96)

1.10 (1.05 to 1.15)

Figure. Cumulative Probability of Dying by Age for Varying Types of Contact With the Legal System

8951

−1737.1

Intercept only log likelihood

No.

−1684.7

Model log likelihood

−0.32 (0.05) [NA]

0.02 (0.30) [−0.57 to 0.61]

American Indian or Alaska Native
(reference: other)

Log (scale)

−0.17 (0.04) [−0.26 to −0.09]

Maleb

Demographic characteristics

0.09 (0.02) [0.05 to 0.14]

Self-rated healthb

General risk factors

No. of arrests

0.02 (0.02) [−0.01 to 0.06]

1.02 (0.99 to 1.06)
0.82 (0.73 to 0.93)

−0.40 (0.18) [−0.75 to −0.05]

0.67 (0.47 to 0.95)

Time ratio (95% CI)

No. of months incarcerated in adult facility

b (SE) [95% CI]

Incarceration in adult facilityb

Key independent variables (before age of 18 y)

Dependent variable: age at death (18-39 y)

Table 2. Parametric Survival Model Regressing Age at Death on Arrest as a Youth and Incarceration
in an Adult Facility as a Youtha

designed for the crucial developmental years of adolescence, where neuronal and social factors

imprisoned in adult facilities. Youths incarcerated in adult facilities experience a system that is not

exposure with health overall, a much smaller segment has focused on such associations for youths

Although a relatively large body of literature highlights the detrimental associations of incarceration

Discussion

0.47-0.95) (eTable 4 in Supplement 1).

Cumulative probability of death

Indicates P < .05.

Following the Akaike information criterion, it was
determined that the parametric survival model
should be estimated using a lognormal distribution
given the distributional properties of age at death.

July 5, 2023

6/10

The cumulative probability of dying by the
corresponding age was estimated using the results
from the model presented in Table 2. Incarcerated <18
y indicates ever spending time in an adult correctional
facility as a youth. CLS indicates criminal legal system.

b

a

Abbreviation: NA, not applicable.

Incarceration of Youths in an Adult Correctional Facility and Risk of Premature Death

correctional facility and an increased risk of death remained virtually identical (TR, 0.67; 95% CI,

JAMA Network Open | Public Health

c..G

Downloaded From: https://jamanetwork.com/ on 08/02/2023

JAMA Network Open. 2023;6(7):e2321805. doi:10.1001/jamanetworkopen.2023.21805 (Reprinted)

born in the US between 1980 and 1984. Fifth, due to the existing literature and the measures

years of age (n = 225), and the current study can be generalized only to the birth cohort of individuals

does not identify the cause of death. Fourth, a limited number of individuals died between 18 and 39

associated with the increased likelihood of early death. Third, due to data limitations, the NLSY97

facilities among youths transferred to such prisons in the analytical sample is a potential factor

mechanisms for the observed patterns. For example, heterogeneity in the experience of adult

observed findings.40 Second, the data do not provide for a nuanced assessment of the potential

aspects of the NLSY97 cohort used to form the analytical sample that were associated with the

incarcerated in adult facilities, at least 5 limitations should temper the findings. First, there could be

Although the current study illustrated an increased likelihood of early mortality for youths

Limitations

youths to adult facilities to avoid the potential lethality of such exposure.

findings observed in the current study highlight the need for rethinking the practice of transferring

association is causal or a function of the risk factors that youths bring with them (or both), the

incarceration process may be amplified by experiences in adult correctional facilities. Whether the

incarceration.38,39 Consequently, the mortality risk factors that youths bring with them to the

health-related risk factors at much higher rates than youths who do not experience

unrelated or tangentially related to the legal system. Youths who are incarcerated tend to possess

Second, the observed association may not be causal but instead indicative of a process

observed here.

facility. Future research should further assess potential mechanisms of the association

Thus, it is conceivable that the associations are exacerbated when experienced by youths in an adult

These associations, however, were examined in prior literature using adult prisoners in adult facilities.

(eg, increased risk for violent victimization, substance use, disease, and harmful behaviors).6,7,9

mechanisms are plausible that align with prior research on the detrimental effects of incarceration

such as general health, being male, and early contact with the legal system. Several potential

correctional facility as a youth may have an association with early mortality beyond other risk factors,

Overall, these results point to a handful of potential explanations. First, any exposure to an adult

eTable 3, and eTable 4 in Supplement 1).

incarceration in juvenile correctional facilities was not associated with early mortality (eAppendix 3,

adult correctional facilities could be associated with detrimental health outcomes because

models further suggest that the circumstances to which youths are exposed during incarceration in

facility as a youth evinced the highest risk for early mortality. The findings of the supplemental

legal system was associated with an increased risk of premature death, being incarcerated in an adult

those without contact with the legal system. Third, while it appears that any formal contact with the

approximately 3 times higher for those who were incarcerated in an adult facility compared with

where the proportion of youths who were estimated to die by the age of 39 years appeared to be

an adult correctional facility and early mortality remained. This association is illustrated in the Figure,

with health, family background, and socioeconomic status, yet the association between exposure to

of exposure. Second, our multivariable models accounted for several general risk factors associated

analytical sample were incarcerated in an adult facility, there was substantial variability in the length

The analyses revealed 3 key findings. First, while a relatively small proportion of youths in the

longitudinal, nationally representative sample of US youths.

national sample, to our knowledge. The current study addressed this gap in the literature using a

has examined the extent to which these experiences may be associated with premature death in a

associated with a variety of detrimental outcomes related to well-being; however, to date, no study

factors associated with the likelihood of early mortality. Experiences of incarceration have been

youths may not only engage in risky and harmful behaviors37 but they may directly experience risk

encounter a system intended exclusively for socially matured individuals. Within such a system,

July 5, 2023

7/10

Incarceration of Youths in an Adult Correctional Facility and Risk of Premature Death

interact to affect personality and behavioral outcomes across the life course.37 Instead, such youths

JAMA Network Open | Public Health

c..G

Downloaded From: https://jamanetwork.com/ on 08/02/2023

JAMA Network Open. 2023;6(7):e2321805. doi:10.1001/jamanetworkopen.2023.21805 (Reprinted)

5. Daza S, Palloni A, Jones J. The consequences of incarceration for mortality in the United States. Demography.
2020;57(2):577-598. doi:10.1007/s13524-020-00869-5

4. Patterson EJ. The dose-response of time served in prison on mortality: New York State, 1989-2003. Am J Public
Health. 2013;103(3):523-528. doi:10.2105/AJPH.2012.301148

3. Sykes BL, Chavez E, Strong J. Mass incarceration and inmate mortality in the United States—death by design?
JAMA Netw Open. 2021;4(12):e2140349. doi:10.1001/jamanetworkopen.2021.40349

2. Massoglia M, Pare PP, Schnittker J, Gagnon A. The relationship between incarceration and premature adult
mortality: gender specific evidence. Soc Sci Res. 2014;46:142-154. doi:10.1016/j.ssresearch.2014.03.002

REFERENCES
1. Bovell-Ammon BJ, Xuan Z, Paasche-Orlow MK, LaRochelle MR. Association of incarceration with mortality by
race from a national longitudinal cohort study. JAMA Netw Open. 2021;4(12):e2133083. doi:10.1001/
jamanetworkopen.2021.33083

Data Sharing Statement: See Supplement 2.

Conflict of Interest Disclosures: None reported.

Statistical analysis: Silver, Nedelec.

Critical revision of the manuscript for important intellectual content: All authors.

Drafting of the manuscript: All authors.

Acquisition, analysis, or interpretation of data: Silver, Nedelec.

Concept and design: All authors.

Author Contributions: Dr Silver had full access to all of the data in the study and takes responsibility for the
integrity of the data and the accuracy of the data analysis.

Author Affiliations: Center for Legal Systems Research, RTI International, Research Triangle Park, North Carolina
(Silver); Department of Sociology, Anthropology, and Criminal Justice, Rutgers University, Camden, New Jersey
(Semenza); New Jersey Gun Violence Research Center, Department of Urban-Global Health, School of Public
Health, Rutgers University, New Brunswick, New Jersey (Semenza); School of Criminal Justice, University of
Cincinnati, Cincinnati, Ohio (Nedelec).

Corresponding Author: Ian A. Silver, PhD, Center for Legal Systems Research, RTI International,
3040 E Cornwallis Rd, Research Triangle Park, NC 27709 (isilver@rti.org).

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2023 Silver IA
et al. JAMA Network Open.

Correction: This article was corrected on July 24, 2023, to fix an error in wording in the Methods section.

Published: July 5, 2023. doi:10.1001/jamanetworkopen.2023.21805

ARTICLE INFORMATION
Accepted for Publication: May 12, 2023.

adult correctional facilities.

factors to ameliorate the potential extralegal harm—including lethal impacts—of placing youths in

pursuing rehabilitation in prison.9,10,18 Prevention and intervention efforts should be directed at

the results emphasize the importance of considering the health-related needs of youths while

increased risk of mortality further illustrates a need for reassessment of this practice. Furthermore,

harmful. The observed association between youth imprisonment in adult correctional facilities and

Our cohort study illustrates that incarcerating youths in adult correctional facilities is potentially very

Conclusions

available in the NLSY97.

association of interest. Future studies should consider additional confounders not measured or

July 5, 2023

8/10

Incarceration of Youths in an Adult Correctional Facility and Risk of Premature Death

available in the NLSY97, the current study adjusted only for mechanisms known to confound the

JAMA Network Open | Public Health

c..G

Downloaded From: https://jamanetwork.com/ on 08/02/2023

JAMA Network Open. 2023;6(7):e2321805. doi:10.1001/jamanetworkopen.2023.21805 (Reprinted)

29. DeSalvo KB, Bloser N, Reynolds K, He J, Muntner P. Mortality prediction with a single general self-rated health
question: a meta-analysis. J Gen Intern Med. 2006;21(3):267-275. doi:10.1111/j.1525-1497.2005.00291.x

28. Rima D, Mukhamadieva G, Alimkulov E, Duzbayeva S, Beaver KM. An analysis of protective and risk factors
associated with no formal contact with the criminal justice system in a nationally representative sample of males
and females. J Crime Justice. 2023;46:267-281. doi:10.1080/0735648X.2022.2105382

27. Bender K. Why do some maltreated youth become juvenile offenders?: a call for further investigation and
adaptation of youth services. Child Youth Serv Rev. 2010;32(3):466-473. doi:10.1016/j.childyouth.2009.10.022

26. Silver IA, Brookstein AJ, D’Amato C. Juvenile incarceration in an adult correctional facility as a risk factor for
adolescent childrearing? J Adolesc. 2023;95(1):56-69. doi:10.1002/jad.12096

25. Bell FC, Miller ML. Life Tables for the United States Social Security Area, 1900-2100. Social Security
Administration, Office of the Chief Actuary; 2005.

24. Moore W, Pedlow S, Krishnamurty P, Wolter K. National Longitudinal Survey of Youth 1997 (NLSY97). National
Opinion Research Center; 2000:254.

23. Kuanliang A, Sorensen JR, Cunningham MD. Juvenile inmates in an adult prison system: rates of disciplinary
misconduct and violence. Crim Justice Behav. 2008;35(9):1186-1201. doi:10.1177/0093854808322744

22. Tie D, Waugh E. Prison Youth Vulnerability Scale: administration and technical manual. Wellington, New
Zealand: Department of Corrections. Accessed December 20, 2022. https://www.corrections.govt.nz/__data/assets/
pdf_file/0015/10671/pyvsmanual.pdf

21. Soler M. Health issues for adolescents in the justice system. J Adolesc Health. 2002;31(6)(suppl):321-333. doi:
10.1016/S1054-139X(02)00494-9

20. Ng IY, Shen X, Sim H, Sarri RC, Stoffregen E, Shook JJ. Incarcerating juveniles in adult prisons as a factor in
depression. Crim Behav Ment Health. 2011;21(1):21-34. doi:10.1002/cbm.783

19. Murrie DC, Henderson CE, Vincent GM, Rockett JL, Mundt C. Psychiatric symptoms among juveniles
incarcerated in adult prison. Psychiatr Serv. 2009;60(8):1092-1097. doi:10.1176/ps.2009.60.8.1092

18. Fagan J, Kupchik A. Juvenile incarceration and the pains of imprisonment. Duke Forum for Law & Social
Change, Fall 2011. Columbia Public Law Research Paper No. 11-263. Posted February 27, 2011. Last revised May 26,
2011. Accessed May 24, 2023. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1772187 doi:10.2139/
ssrn.1772187

17. Puzzanchera C, Adams B, Sickmund M. Juvenile court statistics 2006-2007. National Center for Juvenile
Justice. March 2010. Accessed November 30, 2022. https://www.ojjdp.gov/ojstatbb/njcda/pdf/jcs2007.pdf

16. Kupchik A. The correctional experiences of youth in adult and juvenile prisons. Justice Q. 2007;24(2):247-270.
doi:10.1080/07418820701294805

15. Steinberg L. Adolescent development and juvenile justice. Annu Rev Clin Psychol. 2009;5:459-485. doi:10.
1146/annurev.clinpsy.032408.153603

14. Lambie I, Randell I. The impact of incarceration on juvenile offenders. Clin Psychol Rev. 2013;33(3):448-459.
doi:10.1016/j.cpr.2013.01.007

13. Barnert ES, Perry R, Morris RE. Juvenile incarceration and health. Acad Pediatr. 2016;16(2):99-109. doi:10.
1016/j.acap.2015.09.004

12. Aalsma MC, Lau KSL, Perkins AJ, et al. Mortality of youth offenders along a continuum of justice system
involvement. Am J Prev Med. 2016;50(3):303-310. doi:10.1016/j.amepre.2015.08.030

11. The United States Department of Justice. Justice Manual: Title 9: Criminal: 9-8.000—juveniles. Accessed May
8, 2023. https://www.justice.gov/jm/jm-9-8000-juveniles

10. Pridemore WA. The mortality penalty of incarceration: evidence from a population-based case-control study
of working-age males. J Health Soc Behav. 2014;55(2):215-233. doi:10.1177/0022146514533119

9. Massoglia M, Pridemore WA. Incarceration and health. Annu Rev Sociol. 2015;41:291-310. doi:10.1146/annurevsoc-073014-112326

8. Møller LF, Matic S, van den Bergh BJ, Moloney K, Hayton P, Gatherer A. Acute drug-related mortality of people
recently released from prisons. Public Health. 2010;124(11):637-639. doi:10.1016/j.puhe.2010.08.012

7. Binswanger IA, Blatchford PJ, Lindsay RG, Stern MF. Risk factors for all-cause, overdose and early deaths after
release from prison in Washington state. Drug Alcohol Depend. 2011;117(1):1-6. doi:10.1016/j.drugalcdep.2010.
11.029

July 5, 2023

9/10

Incarceration of Youths in an Adult Correctional Facility and Risk of Premature Death

6. Ruch DA, Steelesmith DL, Brock G, et al. Mortality and cause of death among youths previously incarcerated in
the juvenile legal system. JAMA Netw Open. 2021;4(12):e2140352. doi:10.1001/jamanetworkopen.2021.40352

JAMA Network Open | Public Health

c..G

Downloaded From: https://jamanetwork.com/ on 08/02/2023

JAMA Network Open. 2023;6(7):e2321805. doi:10.1001/jamanetworkopen.2023.21805 (Reprinted)

SUPPLEMENT 2.
Data Sharing Statement

SUPPLEMENT 1.
eAppendix 1. Directed Acyclic Graph to Guide Model Selection
eReferences.
eFigure. Directed Acyclic Graph Guiding the Selection of Covariates
eAppendix 2. Missing Case Information
eTable 1. Descriptive Statistics for the Analytical Sample Before Imputation (N = 8,951)
eTable 2. Missing Case Comparison for Key Constructs Before Imputation (n = 8,951)
eAppendix 3. Model Replications With Youth Incarcerated in Juvenile Facilities
eTable 3. Parametric Survival Model Regressing Age at Death on Being Incarcerated in a Juvenile Facility
eTable 4. Parametric Survival Model Regressing Age at Death on Being Incarcerated in a Juvenile Facility, Arrested
as a Juvenile, and Incarcerated in an Adult Facility as a Juvenile
eAppendix 4. R Script

40. Latzer B. The Rise and Fall of Violent Crime in America. Encounter Books; 2016.

39. Pyle N, Flower A, Fall AM, Williams J. Individual-level risk factors of incarcerated youth. Remedial Spec Educ.
2016;37(3):172-186. doi:10.1177/0741932515593383

38. Forrest CB, Tambor E, Riley AW, Ensminger ME, Starfield B. The health profile of incarcerated male youths.
Pediatrics. 2000;105(1, pt 3)(suppl 2):286-291. doi:10.1542/peds.105.S2.286

37. Moffitt TE, Arseneault L, Belsky D, et al. A gradient of childhood self-control predicts health, wealth, and public
safety. Proc Natl Acad Sci U S A. 2011;108(7):2693-2698. doi:10.1073/pnas.1010076108

36. Zhou H, Wang H, Wang S, Zou Y. SurvMetrics: an R package for predictive evaluation metrics in survival
analysis. The R Journal. Accessed May 24, 2023. https://journal.r-project.org/articles/RJ-2023-009/ doi:10.32614/
RJ-2023-009

35. Therneau T. A package for survival analysis in R. R package version 3.1-11. Accessed October 30, 2022. https://
CRAN.R-project.org/package=survival

34. Ng’andu NH. An empirical comparison of statistical tests for assessing the proportional hazards assumption of
Cox’s model. Stat Med. 1997;16(6):611-626. doi:10.1002/(SICI)1097-0258(19970330)16:6<611::AID-SIM437>3.0.
CO;2-T

33. Kleinbaum DG, Klein M. Evaluating the proportional hazards assumption. In: Survival Analysis: A Self-Learning
Text. 3rd ed. Springer; 2012:161-200.

32. Hess KR. Graphical methods for assessing violations of the proportional hazards assumption in Cox regression.
Stat Med. 1995;14(15):1707-1723. doi:10.1002/sim.4780141510

31. Van Buuren S, Groothuis-Oudshoorn K. mice: Multivariate imputation by chained equations in R. J Stat Softw.
2011;45:1-67. doi:10.18637/jss.v045.i03

July 5, 2023

10/10

Incarceration of Youths in an Adult Correctional Facility and Risk of Premature Death

30. Idler EL, Benyamini Y. Self-rated health and mortality: a review of twenty-seven community studies. J Health
Soc Behav. 1997;38(1):21-37. doi:10.2307/2955359

JAMA Network Open | Public Health

c..G