Confidential. Do not distribute. Pre-embargo material.
ORIGINAL INVESTIGATION

Incarceration, Incident Hypertension,
and Access to Health Care
Findings From the Coronary Artery Risk Development
in Young Adults (CARDIA) Study
Emily A. Wang, MD; Mark Pletcher, MD, MPH; Feng Lin, MS; Eric Vittinghoff, PhD, MPH;
Stefan G. Kertesz, MD, MSc; Catarina I. Kiefe, MD, PhD; Kirsten Bibbins-Domingo, PhD, MD

Background: Incarceration is associated with increased cardiovascular disease mortality, but prospective studies exploring mechanisms of this association are
lacking.
Methods: We examined the independent association of
prior incarceration with incident hypertension, diabetes, and dyslipidemia using the Coronary Artery Risk Development in Young Adults (CARDIA) study—a cohort
of young adults aged 18 to 30 years at enrollment in 19851986, balanced by sex, race (black and white), and education (high school education or less). We also examined the association of incarceration with left ventricular
hypertrophy on echocardiography and with barriers to
health care access.
Results: Of 4350 participants, 288 (7%) reported previous incarceration. Incident hypertension in young adulthood was more common among former inmates than in
those without incarceration history (12% vs 7%; odds ratio, 1.7 [95% confidence interval {CI}, 1.2-2.6]), and this

Author Affiliations: Division of
General Internal Medicine,
San Francisco General Hospital
(Drs Wang and
Bibbins-Domingo), Department
of Medicine (Drs Wang,
Pletcher, and Bibbins-Domingo),
Department of Epidemiology
and Biostatistics (Drs Pletcher,
Vittinghoff, and
Bibbins-Domingo and
Ms Lin), and Center for
Vulnerable Populations
(Dr Bibbins-Domingo),
University of California,
San Francisco; and Division
of Preventive Medicine,
Department of Medicine,
University of Alabama at
Birmingham (Drs Kertesz and
Kiefe). Dr Wang is now with the
Division of General Internal
Medicine, Yale University School
of Medicine, New Haven,
Connecticut.

association persisted after adjustment for smoking, alcohol and illicit drug use, and family income (adjusted
odds ratio [AOR], 1.6 [95% CI, 1.0-2.6]). Incarceration
was significantly associated with incident hypertension
in those groups with the highest prevalence of prior incarceration, ie, black men (AOR, 1.9 [95% CI, 1.1-3.5])
and less-educated participants (AOR, 4.0 [95% CI, 1.017.3]). Former inmates were more likely to have left ventricular hypertrophy (AOR, 2.7, [95% CI, 0.9-7.9]) and
to report no regular source for medical care (AOR, 2.5,
[95% CI, 1.3-4.8]). Cholesterol levels and diabetes rates
did not differ by history of incarceration.
Conclusions: Incarceration is associated with future hypertension and left ventricular hypertrophy among young
adults. Identification and treatment of hypertension may
be important in reducing cardiovascular disease risk
among formerly incarcerated individuals.

Arch Intern Med. 2009;169(7):687-693

I

NCARCERATION HAS BECOME INcreasingly frequent in the lives of
young adults. Between 1987 and
2007, the US prison population
tripled, such that currently 1 in
30 men between the ages of 20 and 34 is
behind bars and 1 in 9 black men in this
age group is incarcerated.1 This rise in incarceration as a normative experience for
young men and young black men in particular makes it especially important to understand the implications of incarceration on future health status.
While the health and health care of prisoners has received some attention, little is
known about the health status of those
with a history of incarceration. One large
study of recently released prisoners from
Washington State demonstrated an increased risk of death immediately following their release from prison.2 The second most common cause of death was from

(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 7), APR 13, 2009
687

cardiovascular disease (CVD), although
the mechanisms of this increased risk were
not examined in this study.
Increases in CVD risk factors associated with incarceration may explain part
of the increased risk of heart disease in prisoners,3 but to our knowledge, no prospective study to date has directly measured
such risk factors in either current or former
US inmates. The Coronary Artery Risk Development in Young Adults (CARDIA) cohort provides the unique opportunity to
explore the development of CVD risk in
young adults with a history of incarceration. The objectives of this study were to
examine prospectively whether a history
of incarceration is associated with the development of CVD risk factors, to explore possible mechanisms and outcomes of this association, and to examine
access to health care among those with
prior incarceration.

WWW.ARCHINTERNMED.COM

©2009 American Medical Association. All rights reserved.

Confidential. Do not distribute. Pre-embargo material.
METHODS

SAMPLE AND SETTING
The CARDIA study is a longitudinal investigation of CVD risk
factors and subclinical coronary disease in a population of black
and white men and women aged 18 to 30 years at baseline in
1985-1986; it is described in detail in previous publications.4
At baseline, the study enrolled 5115 young adults, who were
recruited from 4 US cities (Birmingham, Alabama; Chicago,
Illinois; Minneapolis, Minnesota; and Oakland, California). The
sampling strategy resulted in a cohort balanced by race (52%
black and 48% white), sex (55% male and 45% female), age
(45% aged 18-24 years and 55% aged 25-30 years), and education level (40% with Յ12 years and 60% with Ͼ12 years.)
Subsequent re-examinations have taken place at years 2 (19871988), 5 (1990-1991), 7 (1992-1993), 10 (1995-1996), 15
(2000-2001), and 20 (2005-2006), with high retention rates
(91% at year 2, 86% at year 5, 81% at year 7, 79% at year 10,
73% at year 15, and 69% at year 20.) Of the 5115 participants
in the CARDIA study, only those who had complete baseline
data and follow-up through the year 5 examination were included in the primary analysis (N=4350).

HISTORY OF INCARCERATION
To assess history of incarceration, participants at baseline (19851986) and year 2 (1987-1988) were asked the following question: “During the past year, did any of the following happen to
you?” and “Went to jail” was one of the prespecified responses. In year 2, participants could report any incarceration
event occurring since the prior examination. Responses to these
2 items permitted us to develop an incarceration exposure variable reflecting any jail time during a 3-year period of young
adulthood, extending from 1 year prior to study entry through
the 2 years preceding the year 2 examination. Since most jail
experiences occur during early adult years,1 this measure offers a strong indicator of incarceration in young adulthood.

CVD RISK FACTORS
Cardiovascular disease risk factors were measured at the year
5 examination (1990-1991). Trained and certified technicians
used a random zero sphygmomanometer to record participants’ blood pressure at the year 5 examination and all subsequent CARDIA study examinations. Measurements of systolic
and diastolic blood pressure were taken 3 times at 1-minute
intervals. For the purposes of this analysis, we used a mean of
the second and third blood pressure measurement. Hypertension was defined by a systolic blood pressure of 140 mm Hg or
higher, diastolic blood pressure of 90 mm Hg or higher, or use
of antihypertensive medication. Incident hypertension was defined as meeting these criteria by the year 5 examination in persons without hypertension at baseline. Diabetes was defined
by use of antidiabetic medication. Serum total cholesterol and
high-density lipoprotein cholesterol levels were measured, and
low-density lipoprotein cholesterol level was calculated using
the Friedewald equation.

beam scale, and height was measured to the nearest 0.5 cm using
a wall-mounted stadiometer. Use of cigarettes was ascertained
with interviewer-administered questionnaire. For cigarette smoking, participants were categorized as current smokers if they
had smoked more than 5 cigarettes per week for the past 3
months and former smokers if they had ever smoked.
Use of cocaine, amphetamines, and excessive alcohol were
also ascertained with interviewer-administered questionnaires. For cocaine or amphetamine use, participants were categorized as current users if they had used in the past month,
former users if they had ever used but not in the past month,
or never users. For alcohol use, participants were categorized
as excessive alcohol consumers based on the at-risk consensus thresholds of the National Institute on Alcohol Abuse and
Alcoholism, which were met if a man consumed 14 drinks per
week or more and a woman consumed more than 7 drinks per
week, where a drink was counted as 360 mL of beer, 150 mL
of wine, or 45 mL of spirits.5
Socioeconomic status of each participant was approximated using a measure of family income at year 5. Participants
were asked which category best describes their total combined family income for the past 12 months. This included income from all sources (eg, wages, veteran’s benefits, help from
relatives, and rent from properties [before taxes]). Those who
responded that their family income was less than $ 24 999 were
defined as meeting 200% of the federal poverty line and being
of low socioeconomic status.6

ECHOCARDIOGRAPHY MEASURES
During year 5, all CARDIA study participants underwent 2-dimensional–guided M-mode echocardiography as described previously.7 Left ventricular mass was measured in grams and indexed to body surface area measured as height in meters squared.
We defined left ventricular hypertrophy (LVH) as left ventricular mass index greater than 90 g/m2.

HEALTH CARE ACCESS
To assess health care access barriers, participants were asked
the following 3 questions: (1) “In the past 2 years, have you
always had health insurance or other coverage for medical care?”;
(2) “Do you have a usual source of care? By that we mean
the place you go if you need a checkup or if you are ill”; and
(3) “Was there anytime during the past 2 years when you did
not seek medical care because it was too expensive or health
insurance did not cover it?” Negative answers to question 1 and
2 indicated an insurance barrier and regular care barrier, respectively. An affirmative answer to question 3 was considered an expense barrier.
Realizing that we did not have health care access data at or
before year 7 (1992-1993), we could not perform a formal mediation analysis to test whether health care access might explain any associations of incarceration with CVD risk. Instead, we studied potential associations between a history of
incarceration and subsequent barriers to health care access. If
present, such associations might only intensify the increased
CVD risk of those with a history of incarceration.

STATISTICAL ANALYSIS
POTENTIAL CONFOUNDERS
Body mass index (BMI) (calculated as weight in kilograms divided by height in meters squared) and smoking history were
measured and examined as potential confounders for the association between incarceration and CVD risk factors. Participants were weighed in light clothing using a standard balance

CARDIA study participants with and without a history of incarceration in the 3-year exposure period captured by our data
were first compared for baseline sociodemographic characteristics, CVD risk factors, and potential explanatory risk factors
of disease using the paired t test and ␹2 test as appropriate. The
primary analyses examined the association of incarceration his-

(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 7), APR 13, 2009
688

WWW.ARCHINTERNMED.COM

©2009 American Medical Association. All rights reserved.

Confidential. Do not distribute. Pre-embargo material.
tory in the first 3 years of the CARDIA study and CVD risk factors at the year 5 examination, as well as the incidence of these
risk factors by year 5. We chose year 5 for an assessment of
these outcomes because of the difficulty in postulating the
mechanisms whereby incarceration may contribute to hypertension further in the future.
We next examined the independent association between incarceration and the incidence of these risk factors using multivariable logistic regression and explored covariates of the associations using staged models, first accounting for
demographics, clinical risk factors, and behavioral risk factors
(illicit drug use and excessive alcohol consumption) and then
adding a measure of socioeconomic status. PϽ.05 was considered statistically significant. Because those incarcerated were
mostly male, black, and with limited education (and because
these were sampling strata within the original CARDIA study
design), we repeated our analyses within subgroups defined by
sex, race, and education and tested for interactions between incarceration and these factors. Because of differential dropout
across these strata, we also ran models using inverse probabilityof-censoring weights to reduce bias potentially resulting from
differential dropout.8
In secondary analyses, we examined whether measures of
left ventricular mass index and LVH differed by prior incarceration using Fisher exact, t, and ␹2 tests as appropriate and
multivariable logistic regression models. Finally, among persons with hypertension evident by year 5, we investigated the
association between incarceration and barriers to health care
access in year 7 using multivariable logistic regression.
RESULTS

Of the 4350 participants included in this analysis, 288
(7%) reported having been incarcerated during the
3-year period of young adulthood captured in the first 2
surveys. Black men and less-educated participants were
most likely to have a history of prior incarceration
(Table 1). Within these subgroups, former inmates
were more likely to report family earnings below 200%
of the federal poverty line, smoking, illicit drug use, and
excess alcohol consumption compared with those without incarceration history.
By the year 5 examination, persons with incarceration history had higher mean systolic blood pressures and
were more likely to have hypertension (Table 2). Among
participants without hypertension at baseline, former inmates were more likely to have developed incident hypertension by the year 5 examination than the participants without history of incarceration (12% vs 7%;
unadjusted odds ratio [OR], 1.7 [95% confidence interval {CI}, 1.2-2.6]). Among black men and less-educated
individuals, subgroups of the CARDIA study sample in
whom incarceration was more common, the relationship between incarceration and hypertension was particularly pronounced and statistically significant. Cholesterol levels (low-density lipoprotein cholesterol, 108
mg/dL vs 108 mg/dL [to convert cholesterol to millimoles per liter, multiply by 0.0259] [P = .93]; highdensity lipoprotein cholesterol, 53 mg/dL vs 53 mg/dL
[P=.44]) and diabetes (2% vs 3% [P = .81]) did not differ between participants with and without a history of
incarceration, respectively, even in subgroup analyses.
Because of the association of incarceration with hypertension, we examined several potential explanatory

factors for the observed association of prior incarceration and incident hypertension. Adjustment for age, sex,
race, and drugs and alcohol use did not alter this association (adjusted OR [AOR], 1.6 [95% CI, 1.0-2.5]), nor
did additional adjustment for poverty (AOR, 1.6 [95%
CI, 1.0-2.6]). Although the statistical tests for interaction between incarceration and race, sex, and education
were not significant, we ran these models within the sampling strata of the CARDIA study (Figure). Subgroups
with the highest rate of incarceration—black men and
less-educated participants—showed significant associations between incarceration and incident hypertension
in adjusted models; associations in subgroups that had
lower rates of incarceration were not significant but had
wide confidence intervals. Although black men and lesseducated individuals were more likely to have been lost
to follow-up by year 5, our results were no different in
analyses using inverse probability weights to reduce bias
due to differential dropout (AOR for all participants, 1.5
[95% CI, 0.9-2.2]; black men, 1.9 [95% CI, 1.0-3.6]; and
less-educated participants, 1.6 [95% CI, 0.9-2.7]).
We examined whether incarceration might be associated with end-organ damage related to hypertension. Persons with incarceration history had higher mean (SD) left
ventricular mass index (54.0 [14.1] g/m2 vs 50.3 [12.9] g/m2
[PϽ.001]) and were more likely to have LVH (2% vs 0.6%
[P=.005]). Prior incarceration was associated with LVH in
unadjusted analyses, and we observed a trend for an association even after accounting for potential confounders
(Table 3). In analyses restricted to black men and lesseducated individuals, prior incarceration was also associated with LVH in both unadjusted and adjusted models.
Among participants with hypertension at the year 5
examination, those with a history of incarceration had
an increased odds of reporting a barrier to health care at
the next follow-up visit 2 years later (Table 4). This association persisted after adjustment for age, race, sex, and
socioeconomic status and was more pronounced in black
and less-educated participants. Former inmates were also
more likely to lack treatment for their hypertension at
the year 7 examination (17% [former inmates] vs 41%
[no prior incarceration] treated; unadjusted OR, 3.3 [95%
CI, 1.3-9.1]) and in each of the follow-up visits during
the entire 20-year duration of the CARDIA study (unadjusted OR at year 20, 2.0 [95% CI, 1.3-3.0]). Incarceration prior to 1987 was associated with a 5.7–mm Hg
higher mean systolic blood pressure at the year 20 examination (95% CI, 0.2-11.2 mm Hg), but not a higher
diastolic blood pressure at year 20 (1.1 mm Hg [95% CI,
−0.6 to 2.8 mm Hg]) in analyses adjusting for the age,
sex, and race of these participants.
COMMENT

In a well-characterized US cohort of black and white young
adults, we found that a history of incarceration is associated with a significantly elevated risk of future hypertension and with LVH. During the 3- to 5-year period that
followed incarceration, we found a cumulative incidence of hypertension of 12% among these young adults
aged 23 to 35 years, compared with 7% among those with-

(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 7), APR 13, 2009
689

WWW.ARCHINTERNMED.COM

©2009 American Medical Association. All rights reserved.

Confidential. Do not distribute. Pre-embargo material.
Table 1. Sociodemographic and Clinical Characteristics of 4350 Black and White Men and Women
by History of Incarceration Before 1987 a

Characteristic
All participants (N = 4350)
Age, mean (SD), y
Income Ͻ200% federal poverty line
BMI, mean (SD)
Tobacco use (current or former)
Cocaine use (current or former)
Amphetamine use (current or former)
Alcohol use (excessive)
Black men (n = 905)
Age, mean (SD), y
Income Ͻ200% federal poverty line
BMI, mean (SD)
Tobacco use (current or former)
Cocaine use (current or former)
Amphetamine use (current or former)
Alcohol use (excessive)
White men (n = 1054)
Age, mean (SD), y
Income Ͻ200% federal poverty line
BMI, mean (SD)
Tobacco use (current or former)
Cocaine use (current or former)
Amphetamine use (current or former
Alcohol use (excessive)
Black women (n = 1214)
Age, mean (SD), y
Income Ͻ200% federal poverty line
BMI, mean (SD)
Tobacco use (current or former)
Cocaine use (current or former)
Amphetamine use (current or former)
Alcohol use (excessive)
White women (n = 1177)
Age, mean (SD), y
Income Ͻ200% federal poverty line
BMI, mean (SD)
Tobacco use (current or former)
Cocaine use (current or former)
Amphetamine use (current or former)
Alcohol use (excessive)
ՅHigh school education (n = 1640)
Age, mean (SD), y
Male
Black
Income Ͻ200% federal poverty line
BMI, mean (SD)
Tobacco use (current or former)
Cocaine use (current or former)
Amphetamine use (current or former)
Alcohol use (excessive)
ϾHigh school (n = 2710)
Age, mean (SD), y
Male
Black
Income below 200% federal poverty line
BMI, mean (SD)
Tobacco use (current or former)
Cocaine use (current or former)
Amphetamine use (current or former)
Alcohol use (excessive)

Prior
Incarceration

No Prior
Incarceration

288 (7)
24.0 (3.7)
173 (60)
26 (5)
182 (63)
153 (55)
86 (30)
99 (34)
154 (53)
23.8 (0.3)
104 (68)
25.8 (0.4)
98 (64)
83 (55)
33 (22)
67 (50)
62 (22)
24.2 (0.5)
25 (40)
25.1 (0.6)
39 (63)
39 (63)
34 (55)
23 (40)
49 (17)
23.9 (0.5)
32 (67)
27.7 (1.1)
38 (78)
21 (44)
6 (12)
8 (19)
23 (8)
24.7 (0.8)
12 (52)
25.4 (1.0)
7 (30)
13 (57)
13 (57)
1 (5)
182 (63)
23.3 (0.3)
144 (79)
137 (75)
121 (67)
25.8 (0.4)
129 (71)
96 (54)
49 (27)
72 (45)
106 (37)
25.1 (0.3)
72 (68)
66 (13)
52 (49)
26.3 (0.6)
53 (50)
60 (57)
37 (35)
27 (30)

4062 (93)
25.1 (3.6)
1462 (36)
26 (6)
1269 (31)
1386 (35)
1030 (26)
549 (13)
751 (18)
24.4 (0.1)
317 (43)
26.6 (0.2)
288 (39)
224 (30)
99 (13)
171 (28)
992 (24)
25.6 (0.1)
263 (27)
25.6 (0.1)
296 (30)
448 (45)
377 (38)
210 (23)
1165 (29)
24.5 (0.1)
589 (51)
28.1 (0.2)
388 (33)
238 (21)
125 (11)
8 (1)
1154 (28)
25.6 (0.1)
293 (25)
24.3 (0.2)
297 (26)
472 (41)
429 (38)
94 (9)
1458 (36)
24.1 (0.1)
625 (43)
913 (63)
784 (54)
26.8 (0.2)
660 (45)
477 (33)
336 (23)
256 (22)
2604 (64)
25.6 (0.1)
1118 (43)
1003 (39)
678 (26)
25.8 (0.1)
609 (23)
909 (35)
694 (27)
293 (13)

P Value
Ͻ.001
Ͻ.001
.59
Ͻ.001
Ͻ.001
.05
Ͻ.001
.08
Ͻ.001
.12
Ͻ.001
Ͻ.001
.008
Ͻ.001
.002
.02
.37
Ͻ.001
Ͻ.001
.03
.004
.27
.03
.66
Ͻ.001
Ͻ.001
.74
.03
.004
.45
.36
.61
.24
Ͻ.001
.63
.02
Ͻ.001
.001
.001
.05
Ͻ.001
Ͻ.001
.29
Ͻ.001
.11
Ͻ.001
Ͻ.001
Ͻ.001
.37
Ͻ.001
Ͻ.001
.10
Ͻ.001

Abbreviation: BMI body mass index (calculated as weight in kilograms divided by height in meters squared).
are given as number (percentage) of participants unless otherwise indicated. Percentages are column percentages; for each covariate, the percentages
indicate percentages within strata.
a Data

(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 7), APR 13, 2009
690

WWW.ARCHINTERNMED.COM

©2009 American Medical Association. All rights reserved.

Confidential. Do not distribute. Pre-embargo material.
Table 2. Blood Pressure and Hypertension Among
the CARDIA Study Cohort in 1990-1991 at Age 23 to 35 Years
by History of Incarceration Before 1987

All participants
Black men
White men
Black women

Prior
No Prior
P
Incarceration Incarceration Value
All participants (N = 4350)
SBP, mean (SD), mm Hg
DBP, mean (SD), mm Hg
Hypertension, No. (%)
Incident hypertension, No. (%) a
Black men (n = 905)
SBP, mean (SD), mm Hg
DBP, mean (SD), mm Hg
Hypertension, No. (%)
Incident hypertension, No. (%) a
White men (n = 1054)
SBP, mean (SD), mm Hg
DBP, mean (SD), mm Hg
Hypertension, No. (%)
Incident hypertension, No. (%) a
Black women (n = 1214)
SBP, mean (SD), mm Hg
DBP, mean (SD), mm Hg
Hypertension, No. (%)
Incident hypertension, No. (%) a
White women (n = 1177)
SBP, mean (SD), mm Hg
DBP, mean (SD), mm Hg
Hypertension, No. (%)
Incident hypertension, No. (%) a
ՅHigh school education
(n = 1640)
SBP, mean (SD), mm Hg
DBP mean (SD), mm Hg
Hypertension, No. (%)
Incident hypertension, No. (%) a
ϾHigh school education
(n = 2710)
SBP, mean (SD), mm Hg
DBP, mean (SD), mm Hg
Hypertension, No. (%)
Incident hypertension, No. (%) a

White women
Less educated

111 (13)
70 (11)
65 (23)
29 (12)

108 (11)
69 (10)
650 (16)
254 (7)

Ͻ.001
.42
.006
.007

114 (12)
71 (11)
43 (23)
22 (17)

114 (11)
73 (10)
155 (20)
75 (11)

.73
.10
.05
.08

111 (12)
70 (10)
10 (16)
4 (7)

110 (11)
71 (9)
149 (15)
39 (4)

.57
.33
.81
.35

106 (14)
67 (13)
8 (16)
2 (5)

107 (11)
69 (10)
235 (20)
96 (9)

.39
.09
.51
.29

102 (10)
67 (13)
4 (17)
1 (5)

101 (10)
65 (8)
123 (10)
44 (7)

Ͻ.001
.42
.30
.84

111 (12)
70 (12)
46 (25)
22 (14)

109 (12)
70 (10)
279 (19)
111 (9)

.01
.84
.05
.02

110 (13)
70 (12)
19 (18)
7 (7)

107 (11)
69 (10)
383 (14)
143 (6)

.006
.49
.83
.58

Abbreviations: CARDIA, Coronary Artery Risk Development in Young Adults;
DBP, diastolic blood pressure; SBP, systolic blood pressure.
a A total of 458 participants with baseline hypertension were excluded from
this analysis (n = 3892).

out a history of incarceration. This association was
strongest in those groups most likely to be incarcerated—black men and those with limited education—
and was independent of alcohol and illicit drug use.
Prior incarceration was also significantly associated
with future barriers to health care access. Improving access to health care among former inmates as well as
screening for hypertension and other modifiable hypertension risk factors may be important in reducing the
risk of CVD death among individuals with a history of
incarceration.
The mechanisms by which incarceration may lead to
hypertension are not known. Several mechanisms have
been postulated in prior studies, including higher use of
drugs and alcohol, increased obesity, or lower socioeconomic status.9-12 Our data suggest that these mechanisms do not entirely explain the hypertension risk among
former inmates, since we observed a significant residual

More educated
0.1

1

10

100

Odds Ratio (95% Confidence Interval)

Figure. Association of incarceration history before 1987 with subsequent
incident hypertension (1990-1991) at age 23 to 35 years. Odds ratios were
adjusted for traditional clinical risk factors for hypertension including body
mass index, smoking, excessive alcohol consumption, and illicit drug use,
including cocaine and amphetamine use. The areas under the curve for the
models including all participants ranged from 0.52 to 0.68. The interaction
terms between incarceration and race (P = .77), incarceration and sex
(P = .15), and incarceration and educational level (P = .22) were not
statistically significant.

effect even after accounting for these factors. Other possible explanations include increased hostility or stress
among individuals with prior incarceration that has
been shown to increase the risk for hypertension and
ultimately atherosclerosis.13-15 The stress of incarceration may increase catecholamine or stress hormone
levels that lead to hypertension, or incarceration may
cause lasting dysregulation of these hormones that
might lead to the development of hypertension later in
life at faster rates.
Although exploring these mechanisms whereby incarceration may contribute to hypertension is beyond the
scope of the present study, our results suggest that hypertension and associated LVH in young former inmates may contribute to the previously observed increased risk of CVD death after their release. Left
ventricular hypertrophy is a well-established independent risk factor for the development of heart failure and
mortality.16,17 Further study is needed to explore this relationship between incarceration, hypertension, and LVH,
both to establish whether the association is causal—and
if so, to elucidate how a history of incarceration might
lead to hypertension and cardiovascular damage—and
whether this is related to the type and duration of incarceration exposure.
While incarceration is not a traditional risk factor for
CVD, our results suggest that a history of incarceration
should be understood as part of the risk profile for the
development of hypertension and LVH in young adults.
Physicians working in communities where incarceration is highly prevalent should consider screening for a
history of incarceration because it may provide information about the future risk of hypertension and associated end-organ damage, as well as the risk of discontinuity in health care or medical treatment. Moreover,
detention in jail, where health care is constitutionally guaranteed, may present a prime opportunity to screen soonto-be released inmates for hypertension and to link inmates with chronic conditions to health care services in
the community on release.
We find that hypertensive individuals with a history
of incarceration were less likely to have insurance, ac-

(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 7), APR 13, 2009
691

WWW.ARCHINTERNMED.COM

©2009 American Medical Association. All rights reserved.

Confidential. Do not distribute. Pre-embargo material.
Table 3. Association of Incarceration History Before 1987 With Left Ventricular Hypertrophy in 1990-1991 at Age 23 to 35 Years a

Unadjusted

Adjusted for Demographic
Characteristics, Clinical Risk
Factors, and Illicit Drug Use b

Adjusted for Demographic
Characteristics, Clinical Risk
Factors, Illicit Drug Use,
and Socioeconomic Status c

Participants

OR (95% CI)

P Value

OR (95% CI)

P Value

OR (95% CI)

P Value

All participants (n =3688)
Black men (n = 760)
Participants with Յhigh
school education (n=1364)

3.8 (1.4-10.1)
2.6 (0.6-10.6)
4.8 (1.6-14.4)

.009
.18
.006

2.5 (0.8-7.3)
4.3 (0.9-20.0)
4.3 (1.0-17.3)

.14
.07
.04

2.7 (0.9-7.9)
6.0 (1.2-31.0)
4.3 (1.0-17.3)

.08
.03
.04

Abbreviations: CI, confidence interval; OR, odds ratio.
a This analysis included only participants who were not hypertensive at baseline. The areas under the curve for models including all participants ranged from
0.69 to 0.82.
b Traditional clinical risk factors for hypertension include body mass index, smoking, and excessive alcohol consumption; illicit drug use includes cocaine and
amphetamine use.
c Socioeconomic status is defined by family income.

Table 4. Association of Incarceration History With Subsequent Access to Care
Among the 650 CARDIA Study Participants With Hypertension
No Regular Source of Care
Participants
All participants (n =650)
Black men (n = 170)
Participants with Յhigh
school education (n=287)

No Health Insurance

Limited Medical Care

AOR (95% CI) a

P Value

AOR (95% CI)

P Value

AOR (95% CI)

P Value

2.5 (1.3-4.8)
2.9 (1.2-6.6)
2.5 (1.1-5.5)

.005
.01
.03

2.5 (1.4-4.7)
2.4 (1.1-5.4)
3.3 (1.5-6.9)

.003
.03
.02

4.3 (2.1-8.7)
6.5 (2.2-18.8)
5.1 (2.0-13.2)

Ͻ.001
Ͻ.001
Ͻ.001

Abbreviations: AOR, adjusted odds ratio; CARDIA, Coronary Artery Risk Development in Young Adults; CI, confidence interval.
a Adjusted for sex, race, age, and socioeconomic status.

cess to health care, or be using antihypertensive medications. Current and former inmates with chronic diseases such as hypertension typically fall at the intersection
of 2 poorly functioning health care systems: the correctional health care system and the public safety net health
care system. Neither health care system is well equipped
to take care of the growing population of individuals with
chronic medical conditions who cycle in and out of both
systems. Ninety percent of those released from jail are
uninsured and lack financial resources to pay for their
medical care in the community.18 Most state correctional systems do not provide discharged inmates with
a state identification card, thereby rendering such individuals ineligible for care in the major county hospitals
that exist to serve indigent persons on release from correctional facilities.19 With access to regular outpatient care
sharply constrained, formerly incarcerated individuals are
more likely to seek care in the emergency department
rather than a primary care office.19,20 The resulting discontinuity and irregularity of service may lead to poor
health outcomes and/or duplication of services. Improved continuity of care between correctional facilities
and the community could protect the health of inmates
with hypertension and other chronic conditions, but assuring such continuity would require reconfiguration of
correctional services and modest interventions such as
the provision of referrals and identification cards to speed
re-entry.
There are several limitations in our study. History of
incarceration was only measured in the first 2 CARDIA

study examinations, and the single binary question item
captured the preceding year events that would have included both a brief jail stay after arrest as well as longer
periods of imprisonment. Because the CARDIA study
questionnaire asked specifically about jail, we can also
not exclude the possibility that those in prisons would
not have answered affirmatively. We therefore lack accurate information about the duration, frequency, or nature of the incarceration exposure or exposure that might
have occurred well before the baseline examination. Such
information would be useful for understanding the true
effect of the type, frequency, and intensity of incarceration on the development of hypertension, especially given
the high rates of recidivism nationwide.21 Measurement of important confounders in this study, including illicit drug use, was based on self-report. There
may have been social desirability and recall bias at
play in this study, especially in the reporting of illicit
drugs.22,23 Nonetheless, the prevalence of illicit drug
use observed in this study approximates that in other
studies in both incarcerated and nonincarcerated
populations.23,24 Finally, we initially chose to examine
3 types of CVD risk factors and only found an association of incarceration with hypertension, thus raising
the possibility of type 1 error. However, the strength
of this analysis is the consistency of the association we
observe across the population subgroups in the United
States at highest risk of incarceration, as well as the
association of incarceration with the end-organ manifestations of hypertension.

(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 7), APR 13, 2009
692

WWW.ARCHINTERNMED.COM

©2009 American Medical Association. All rights reserved.

Confidential. Do not distribute. Pre-embargo material.
In conclusion, we found that prior incarceration predicts future hypertension and LVH among young adults.
For the more than 7 million people that pass through US
jails and prisons each year,25 incarceration may be an independent risk factor for the development of hypertension and LVH, both of which put such persons at higher
risk for clinical CVD. Incarceration may be a cause for
hypertension and CVD, but may also present an underused opportunity for intervention and improving health
and access to health care.
Accepted for Publication: November 14, 2008.
Correspondence: Kirsten Bibbins-Domingo, PhD, MD,
Division of General Internal Medicine, Department of
Medicine, University of California, San Francisco, Box
1364, SFGH Bldg 10, WD 13 1313, San Francisco, CA
94143-1364 (bibbinsk@medicine.ucsf.edu).
Author Contributions: Drs Bibbins-Domingo and Wang
had full access to all of the data in the study and take responsibility for the integrity of the data and accuracy of
the data analysis. Study concept and design: Wang, Pletcher,
and Bibbins-Domingo. Acquisition of data: BibbinsDomingo. Analysis and interpretation of data: Wang,
Pletcher, Lin, Vittinghoff, Kertesz, Kiefe, and BibbinsDomingo. Drafting of the manuscript: Wang, Kiefe, and
Bibbins-Domingo. Critical revision of the manuscript for
important intellectual content: Wang, Pletcher, Lin, Vittinghoff, Kertesz (social and policy implications), Kiefe,
and Bibbins-Domingo. Statistical analysis: Wang, Pletcher,
Lin, Vittinghoff, and Kiefe. Administrative, technical, and
material support: Kiefe and Bibbins-Domingo. Study supervision: Bibbins-Domingo.
Financial Disclosure: None reported.
Funding/Support: Work on this manuscript was supported (or partially supported) by the following contracts: University of Alabama at Birmingham, Coordinating Center, grant N01-HC-95095; University of
Alabama at Birmingham, Field Center, grant N01-HC48047; University of Minnesota, Field Center and Diet
Reading Center (year 20 examination), grant N01-HC48048; Northwestern University, Field Center, grant N01HC-48049; Kaiser Foundation Research Institute, grant
N01-HC-48050; University of California, Irvine, Echocardiography Reading Center (years 5 and 10), grant N01HC-45134; Harbor-UCLA Research Education Institute, Computed Tomography Reading Center (year 15
examination), grant N01-HC-05187; Wake Forest University (year 20 examination), grant N01-HC-45205;
and New England Medical Center (year 20 examination), grant N01-HC-45204 from the National Heart, Lung
and Blood Institute. Dr Wang was supported by a National Research Service Award Training Grant in General Internal Medicine to the University of California, San
Francisco (UCSF) (grant T32 HP19025). Dr BibbinsDomingo is supported by grants from the Robert Wood
Johnson Amos Faculty Development Program, a diversity supplement to the CARDIA study contract to the University of Alabama Coordinating Center (grant N01-HC95095), and by a UCSF Hellman Family Faculty Award.
Additional Contributions: Tekeshe Mekonnen, MS,
provided administrative assistance in the resubmission
of the manuscript.

REFERENCES
1. The Pew Charitable Trusts. One in 100: Behind Bars in America 2008. Washington, DC: The Pew Charitable Trusts; 2008.
2. 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.
3. Olubodun J. Prison life and the blood pressure of the inmates of a developing
community prison. J Hum Hypertens. 1996;10(4):235-238.
4. Cutter GR, Burke GL, Dyer AR, et al. Cardiovascular risk factors in young adults:
the CARDIA baseline monograph. Control Clin Trials. 1991;12(1)(suppl):1S77S.
5. National Institute on Alcohol Abuse and Alcoholism. Helping patients who drink
too much: a clinician’s guide. 2008. http://pubs.niaaa.nih.gov/publications
/Practitioner/CliniciansGuide2005/clinicians_guide.htm. Accessed October 15,
2008.
6. US Department of Health and Human Services. Prior HHS poverty guidelines and
Federal Register references. http://aspe.hhs.gov/POVERTY/figures-fed-reg
.shtml. Accessed October 15, 2008.
7. Gardin JM, Wong ND, Bommer W, et al. Echocardiographic design of a multicenter investigation of free-living elderly subjects: the Cardiovascular Health Study.
J Am Soc Echocardiogr. 1992;5(1):63-72.
8. Alonso A, Segui-Gomez M, de Irala J, Sanchez-Villegas A, Beunza JJ, MartinezGonzalez MA. Predictors of follow-up and assessment of selection bias from dropouts using inverse probability weighting in a cohort of university graduates. Eur
J Epidemiol. 2006;21(5):351-358.
9. Hajjar I, Kotchen TA. Trends in prevalence, awareness, treatment, and control of
hypertension in the United States, 1988-2000. JAMA. 2003;290(2):199-206.
10. Centers for Disease Control and Prevention (CDC). State-specific trends in selfreported blood pressure screening and high blood pressure—United States,
1991-1999. MMWR Morb Mortal Wkly Rep. 2002;51(21):456-460.
11. Bolen JC, Rhodes L, Powell-Griner EE, Bland SD, Holtzman D. State-specific prevalence of selected health behaviors, by race and ethnicity–Behavioral Risk Factor
Surveillance System, 1997. MMWR CDC Surveill Summ. 2000;49(2):1-60.
12. Office of National Drug Control Policy. Drug data summary. http://www
.whitehousedrugpolicy.gov/publications/factsht/drugdata/index.html. Accessed April 20, 2008.
13. Yan LL, Liu K, Matthews KA, Daviglus ML, Ferguson TF, Kiefe CI. Psychosocial
factors and risk of hypertension: the Coronary Artery Risk Development in Young
Adults (CARDIA) study. JAMA. 2003;290(16):2138-2148.
14. Carroll D, Davey Smith G, Sheffield D, Shipley MJ, Marmot MG. The relationship
between socioeconomic status, hostility, and blood pressure reactions to mental stress in men: data from the Whitehall II study. Health Psychol. 1997;16
(2):131-136.
15. Iribarren C, Sidney S, Bild DE, et al. Association of hostility with coronary artery
calcification in young adults: the CARDIA study: Coronary Artery Risk Development in Young Adults. JAMA. 2000;283(19):2546-2551.
16. Koren MJDR, Casale PN, Savage DD, Laragh JH. Relation of left ventricular mass
and geometry to morbidity and mortality in uncomplicated essential hypertension.
Ann Intern Med. 1991;114(5):345-352.
17. Levy D, Savage DD, Kannel WB, Castelli WP. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study.
N Engl J Med. 1990;322(22):1561-1566.
18. Lee J, Vlahov D, Freudenberg N. Primary care and health insurance among women
released from New York City jails. J Health Care Poor Underserved. 2006;17
(1):200-217.
19. Mallik-Kane K. Returning Home Illinois Policy Brief: Health and Prison Reentry.
Washington, DC: Urban Institute Justice Policy Center; 2005.
20. Conklin TJ, Lincoln T, Tuthill RW. Self-reported health and prior health behaviors of newly admitted correctional inmates. Am J Public Health. 2000;90(12):
1939-1941.
21. Greenfeld LA. Prison sentence and time served for violence. Washington, DC:
Bureau of Justice Statistics. April 1995. Special Findings Publication NCJ153858.
22. Kim MT, Hill MN. Validity of self-report of illicit drug use in young hypertensive
urban African-American males. Addict Behav. 2003;28(4):795-802.
23. Hill MN, Bone LR, Kim MT, Miller DJ, Dennison CR, Levine DM. Barriers to hypertension care and control in young urban black men. Am J Hypertens. 1999;
12(10 Pt 1):951-958.
24. Mumola C. Substance abuse and treatment, state and federal prisoners, 1997.
Washington, DC: Bureau of Justice Statistics. January 1997. Special Report No.
NCJ-172871.
25. Sabol WJ, Minton TD, Harrison PM. Prison and jail inmates at midyear 2006.
Washington, DC: Bureau of Justice Statistics; June 2006. Bulletin No. NCJ217675.

(REPRINTED) ARCH INTERN MED/ VOL 169 (NO. 7), APR 13, 2009
693

WWW.ARCHINTERNMED.COM

©2009 American Medical Association. All rights reserved.