Content uploaded by Holly Ventura Miller
Author content
All content in this area was uploaded by Holly Ventura Miller on Jun 22, 2017
Content may be subject to copyright.
Am J Health Behav.™™
™™
™ 2011;35(1):15-27 15
Self-control and Health Outcomes in a
Nationally Representative Sample
Holly Ventura Miller, PhD; J.C. Barnes, PhD; Kevin M. Beaver, PhD
Holly Ventura Miller, Assistant Professor, De-
partment of Criminal Justice, University of Texas
at San Antonio, San Antonio, TX. J.C. Barnes,
Assistant Professor, University of Texas at Dal-
las, School of Economic, Political and Policy Sci-
ences, Richardson, TX. Kevin M. Beaver, Assis-
tant Professor, College of Criminology and Crimi-
nal Justice, Florida State University, Tallahas-
see, FL.
Address correspondence to Dr Miller, Depart-
ment of Criminal Justice, University of Texas at
San Antonio, 501 W Durango Blvd, San Antonio,
TX 78207. E-mail: Holly.Miller@utsa.edu
Objectives: To explore the link
between low self-control during ado-
lescence and health problems in
early adulthood. Methods: Using
data from the National Longitudi-
nal Study of Adolescent Health, we
examined the relationship between
varying levels of self-control and
the likelihood of being diagnosed
with a variety of physical and brain-
based health conditions. Results:
Results from logistic regression
analyses indicated that subjects
with lower levels of self-control
had significantly higher odds of
being diagnosed with 9 of the 10
health outcomes. Conclusions: Tar-
geting the development of self-
control in childhood and adoles-
cence may be valuable in prevent-
ing future health problems.
Key words: self-control, health
problems, brain-based disorders
Am J Health Behav. 2011;35(1):15-27
The relationship between self-control
and crime has been well established
in the criminological literature.1,2
Theoretically, however, self-control is also
linked to a range of problem behaviors
and negative outcomes beyond those of
crime and delinquency.3 To date, only a
limited body of empirical research has
examined the association of self-control
with outcomes other than crime and de-
linquency,3-5 though many have long con-
sidered the possibility of a general pro-
pensity toward problem behavior.6-9 Col-
lectively, these studies support the no-
tion of a general cause of, or at least a
general propensity toward, problem be-
haviors and habits such as heavy drink-
ing and drug use, smoking, early sexual
activity, and dangerous driving, among
others.
What has been explored to a lesser
extent is whether a relationship exists
between a well-known predictor of adoles-
cent problem behavior, self-control, and
health outcomes. Because self-control
theory1 explicitly argues that a lack of
self-control can lead to any number of
deleterious consequences, it is plausible
that low self-control may be associated
with experiencing adverse health condi-
tions. We explore this possibility by exam-
ining the relationship between self-con-
trol and a variety of physical (ie, asthma,
cancer, diabetes, high cholesterol, and
high blood pressure) and brain-based (ie,
depression, ADHD, mental illness, poor
hearing, and stuttering) health condi-
tions among a nationally representative
sample of Americans.
The Generality of Low Self-control
Though much of the research is corre-
lational in nature, considerable evidence
exists suggesting that some individuals
possess a general tendency toward prob-
lem behavior. The co-occurrence of many
forms of problem behavior such as drunk
Self-control and Health Outcomes
16
driving, early onset of smoking, use of
illicit drugs, and involvement in other
types of delinquency has been observed
among certain adolescents.7,10,11 Based on
this research, Gottfredson and Hirschi
introduced A General Theory of Crime in
which low self-control was implicated as
the main cause of all deviant or problem
behavior.12 Specifically, they stated that
self-control and involvement in deviant
behavior have an inverse relationship;
those with relatively high levels of self-
control will not succumb to the tempta-
tion of immediate pleasures as easily as
will those with relatively low levels of self-
control. According to the theory, low self-
control is a trait that consists of 6 dimen-
sions: impulsivity, preference for simple
over complex tasks, high propensity for
risk seeking, preference of physical as
opposed to mental activities, self-
centeredness, and possession of a short
temper. In short, individuals possessing
these traits, given the opportunity, will be
more likely to act on their impulses,
deviant or otherwise. Additionally, those
with low self-control will engage in “analo-
gous behaviors” that characterize their
tendency “to pursue immediate pleasures
that are not criminal: they will tend to
smoke, drink, use drugs, gamble, have
children out of wedlock, and engage in
illicit sex.”12
Early research prioritized examining
the theory’s key causal hypothesis: that
self-control has a negative effect on devi-
ance. An impressive line of research has
tested this hypothesis, the results of which
have overwhelmingly shown that low lev-
els of self-control are predictive of a range
of deviant behaviors such as drunken
driving13 and delinquency.14 Overall, re-
search indicates that self-control is one
of the most consistent correlates of devi-
ant or problem behavior in both adoles-
cence and adulthood.
Though few studies have been con-
ducted in which the health-problem be-
havior relationship is explored, those that
do exist confirm that offenders typically
are of poorer health than non-offenders.
Early studies identified constitutional fac-
tors of children, such as health, as corre-
lated with delinquency;17-19 and more re-
cent work has also supported this rela-
tionship.20 Farrington11 also explored the
link between offending and poor health
and found that convicted males tended to
have myriad health-related problems,
from serious illnesses and accidents, to
hospital treatment and road accident in-
juries. Using the same data, Shepherd
and colleagues15 found that childhood pre-
dictors of adolescent offending also pre-
dicted injury, cardiovascular illness, and
psychological problems at ages 27 to 32.
More recently, Samuelson and colleagues
found that adolescent problem or antiso-
cial behavior was predictive of adverse
health outcomes to age 50, including
death, hospitalization, substance misuse,
and mental illness.21 These findings al-
lude to the current study’s hypothesis
that the same underlying cause of devi-
ant behavior (eg, low levels of self-control)
may also be associated with adverse
health conditions.
Self-control Beyond Crime and
Delinquency
As noted, the extant literature indi-
cates that offenders are at higher risk for
suffering from adverse health outcomes
when compared to non-offenders.11,15,17-20
At the same time, researchers have found
self-control to be a robust predictor of
involvement in various types of antiso-
cial, deviant, or problem behavior. Fur-
ther, research has provided support for a
causal link between low self-control and
various negative outcomes such as acci-
dents 4 and drug use and abuse.1,22 Despite
these associations, little research has
examined the relationship between self-
control and health outcomes. Theoreti-
cally, Gottfredson and Hirschi’s theory
provides a useful framework for under-
standing variations in health conditions,
particularly those that may be partially
attributable to lifestyle choices and influ-
ences. Recall that Gottfredson and
Hirschi12 stated that low self-control leads
to “analogous behaviors” such as smok-
ing and illicit sex, behaviors that may be
risk factors for health-related problems.
Research has supported this analogous-
behaviors hypothesis by revealing that
those with low levels of self-control fre-
quently violate normative standards of
behavior, thus placing themselves in situ-
ations or contexts of additional risk of
injury or death.23 It is sensible, then, to
hypothesize that low self-control may also
be associated with experiencing adverse
health conditions.
This theoretical linkage has also been
inferred in the psychological literature.
Self-regulation, for example, like self-
Miller et al
Am J Health Behav.™™
™™
™ 2011;35(1):15-27 17
control, is viewed as a durable and central
aspect of personality that develops early
in life.24,25 Problems with self-regulation
have been linked with the management
of a range of health behaviors and out-
comes including exercise, dieting, smok-
ing cessation, problem drinking, and eat-
ing disorders.25-28,30,31 Those who lack self-
regulation experience difficulties with
self-monitoring, proximal goal setting,
strategy development, and self-motiva-
tion,32 all of which can be both theoreti-
cally and empirically linked to health
functioning.
Albert Bandura’s 1997 work, Self-effi-
cacy: The Exercise of Control, offers a use-
ful framework for linking self-control and
health functioning.32 Although genetics
play an integral role in the etiology of
health conditions and quality, physical
health is also largely determined by
lifestyle habits and environmental condi-
tions. Health habits in particular can
exert a major impact on the quality of both
psychological and physical well-being.32(p257)
In that self-control/regulation is impor-
tant to understanding lifestyle habits, it
is sensible to link self-control with health
conditions via the mechanism of these
habits.
Empirical evidence also appears to sup-
port the link between self-control and
health. In a recent study, Schroder and
Schwarzer31 employed data from 381 heart
patients to test whether habitual self-
control predicted health behaviors follow-
ing surgery. Self-control was assessed
with the 10-item habitual self-control
(HSC) questionnaire developed by
Schroeder.33 Sample items include “If
something important to me turns out to
be quite difficult, I just persist in my
efforts”; “I often have difficulty rejecting a
tempting offer”; “I often find it hard to
bring myself to take an unpleasant but
necessary action.” The authors argued
that self-control can affect health out-
comes by promoting healthy behaviors.
Findings suggested that compared to other
trait variables, self-control is a salient
predictor of health-behavioral outcomes.
Self-control was also able to explain
unique variation in dieting and physical
exercise beyond proximal behavior-spe-
cific predictors (ie, self-efficacy, inten-
tions).
The current study tests the hypothesis
that a known correlate of problem behav-
ior, in this case self-control, may also be
associated with health problems in early
adulthood. Specifically, we argue that it
need not require a necessarily antisocial
lifestyle to predict the development of
adverse health conditions, but rather a
lifestyle distinguished by impulsivity and
poor decision making, behaviors com-
mon to those with low self-control. The
impulsivity and inability to delay gratifi-
cation that characterize the decision-
making processes of those with low self-
control may impact the likelihood of de-
veloping adverse health conditions later
in life. Because health problems are often
the result of long-term negative behav-
iors and habits, it is reasonable to expect
that low levels of self-control would be
correlated with their manifestation. Fur-
thermore, if health-related problems can
be viewed as another form of deviant
behavior,9 or at least the result of deviant
behavior, then self-control may indeed be
associated with the likelihood of occur-
rence.
Although it is fairly well established
that offenders and those who exhibit an-
tisocial or problem behaviors in adoles-
cence are more likely to suffer from del-
eterious health outcomes, this relation-
ship has yet to be situated within a spe-
cific explanatory framework. Using data
from the National Longitudinal Study of
Adolescent Health (Add Health), we as-
sess the relationship between self-con-
trol and adverse health conditions during
early adulthood. We argue that the same
latent personality trait that is known to
predict involvement in problem behavior
(self-control) should also be associated
with health outcomes such as high blood
pressure and depression.
METHODS
Data
The data used for this study come from
the National Longitudinal Study of Ado-
lescent Health (Add Health).34 The Add
Health is a longitudinal and nationally
representative sample of adolescents
enrolled in US middle and high schools.
Sample schools were selected by employ-
ing stratified random sampling techniques
that ultimately resulted in 80 high schools
and 52 middle schools being included.
Between September 1994 and April 1995,
nearly 90,000 students in grades 7
through 12 completed wave I in-school
self-report surveys that asked a variety of
questions about their families, their daily
Self-control and Health Outcomes
18
activities, and their peer relationships.
Also during 1995, a subsample of this
original cohort was chosen to be re-inter-
viewed in their homes, along with their
primary caregiver (typically their mother).
Data were gathered from the primary
caregiver at wave I only. The wave I in-
home surveys were designed to capture
more detailed information about the ado-
lescents such as their social relation-
ships, their behaviors, and the status of
their health. Questions were also asked
that measured psychological character-
istics, including self-control and depres-
sion. These interviews used a mixture of
computer-assisted personal interviews
and audio computer-assisted self-inter-
views. Overall, 20,745 adolescents and
17,700 of their primary caregivers par-
ticipated in the wave I in-home compo-
nent of the study.35 The respondents
ranged between 12 and 21 years at wave
I, with the mean age being 16 years.
Approximately one year after the first
in-home survey was conducted, 14,738 of
the original respondents were interviewed
a second time. Respondents who were in
the 12th grade at wave I were not re-
interviewed at wave II. The wave II in-
home questionnaires were designed to be
very similar to those used at wave I.
Adolescents were again asked about their
peers, their psychological characteris-
tics, and their health. Primary caregivers
were not interviewed at wave II. Finally,
in 2001-2002, approximately 7 years after
the initial interviews were carried out, a
third wave of in-home interviews was
conducted. At this point, most respon-
dents had reached early adulthood and
thus were no longer adolescents. Respon-
dents ranged between 18 and 26 years at
wave III. As a result, the wave III surveys
were altered so that more age-appropri-
ate questions were included on the sur-
vey instruments. Respondents, for ex-
ample, were asked about their employ-
ment history, their involvement in crimi-
nal behaviors, and whether or not they
had been diagnosed with a range of physi-
cal and mental ailments. Of the original
wave I in-home cohort, 15,197 respon-
dents were successfully interviewed at
wave III. After deleting cases with miss-
ing values, the final analytical sample
size ranged between N = 11,955 and N =
12,077. For additional details about the
Add Health respondents or the sampling
design, see Harris et al.35
Measures
Health-related outcomes. During the
wave III interviews, detailed information
about health-related issues was collected
from each respondent. Specifically, a se-
ries of questions were included that ad-
dressed whether or not the respondent
had ever been diagnosed with the follow-
ing 5 health-related outcomes: asthma,
cancer, diabetes, high blood pressure,
and high cholesterol. Questions were coded
dichotomously so that respondents who
had ever been diagnosed with the condi-
tion were assigned a value of “1” and
respondents who had not been diagnosed
with the condition were assigned a value
of “0.” Table 1 presents descriptive values
of the health-problems variables as well
as the other variables/scales used in the
analyses. A complete listing of the exact
questionnaire items used for all vari-
ables is available in Appendix A.
Brain-based disorders. During wave
III interviews, respondents were also
asked a series of questions regarding the
presence of brain-based disorders. Spe-
cifically, respondents were asked
whether they had ever been diagnosed
with depression, whether they had been
prescribed ADD or ADHD medication in
the past 12 months, whether they had
been treated for a mental illness in the
past 5 years, and whether or not they had
problems with stuttering when speaking.
In each case, an affirmative response
was coded as “1” and all others were coded
as “0.” Additionally, respondents were
asked about their hearing ability. Re-
spondents who indicated that their hear-
ing was excellent, good, or fair were coded
as “0,” whereas respondents indicating
that their hearing ability was poor, very
poor, or deaf were coded as “1.” Thus, a
response of “1” to any of these 5 questions
indicated the presence of that condition,
or in the case of the hearing question, the
presence of poor hearing.
Low self-control. There has been con-
siderable debate concerning the most
reliable and valid way to measure levels of
self-control.36-39 Prior researchers analyz-
ing the Add Health data have identified a
5-item low self-control scale.40 We used
these 5 items as a starting point to de-
velop an expanded low self-control scale.
In total, we identified 23 items that were
indicators of low self-control. Items in-
cluded in this scale tapped the
respondent’s ability to plan ahead, tem-
Miller et al
Am J Health Behav.™™
™™
™ 2011;35(1):15-27 19
per, and ability to get along well with
others. Both behavioral and attitudinal
measures were included. For example,
respondents were asked how they re-
acted, behaviorally, to various situations.
Respondents were also asked how they
felt towards themselves and others. It
should also be noted that the scale in-
cludes items reported on by the respon-
dent as well as items reported on by the
respondent’s parent. Exploratory factor
analysis indicated that a single-factor
solution best explained the relationship
among the 23 items. Thus, responses to
each item were summed to form the wave
I low self-control scale. Higher values on
this scale indicated lower levels of self-
control (α = .75).
Polydrug use. During wave III inter-
views, respondents were asked 6 ques-
tions that tapped whether or not they had
used alcohol, marijuana, cocaine, crystal
methamphetamine, drugs such as hallu-
cinogens and illegal prescription medica-
tions, and intravenous drugs in the past
year. Individual items were coded so that
“1” indicated the respondent had used the
drug in the past year and “0” indicated
that the respondent had not used the drug
in the past year. The polydrug use index
was created by summing the responses to
these 6 questions. Thus, the minimum
achievable score on the polydrug use in-
dex was 0, indicating no drug use in the
past year, and the maximum achievable
score was 6, indicating the use of various
drugs in the past year (α = .59).
Demographic variables. To help pre-
vent model misspecification, we included
age, gender, and race as control vari-
ables. Age was a continuous variable
measured in years, whereas gender (0=fe-
male, 1=male) and race (0=white, 1=non-
white) were dichotomous dummy vari-
ables.
Analysis
The analysis proceeded in a 2-step
approach. First, we estimated a series of
binary logistic regression models to ex-
amine the effects of self-control, along
with control variables, on each of the 5
health-related outcomes (ie, asthma, can-
cer, diabetes, high cholesterol, and high
blood pressure). Second, these same mod-
els were estimated using the 5 brain-
based disorders (ie, depression, ADHD,
mental illnesses, poor hearing, and stut-
tering) as dependent variables. All of the
models included the polydrug use index
as a control variable to take account of the
possibility that individuals with lower lev-
els of self-control may be more likely to
use drugs, and thus more likely to experi-
ence adverse health outcomes. In effect,
the polydrug use index is expected to
mediate, at least partially, the effects of
self-control on the outcome variables. All
analyses were conducted using the sta-
tistical package Stata 10.1.
RESULTS
Table 2 presents the logistic regres-
sion models that estimated the odds of
having asthma, cancer, diabetes, high
cholesterol, and high blood pressure. As
can be seen, respondents with lower lev-
els of self-control had significantly higher
odds of being diagnosed with 4 of the 5
conditions: asthma, cancer, high choles-
terol, and high blood pressure. These re-
lationships are most clearly evidenced
by the odds ratios of the low self-control
Table 1
Descriptive Statistics for Add
Health Study Variables
Frequency %
Respondents with…(Wave 3)
Asthma 2,553 16.8
Cancer 13 9 .9
Diabetes 15 2 1. 0
High cholesterol 67 2 4. 4
High blood pressure 846 5 .6
Respondents with…(Wave 3)
Depression 1,586 10.5
ADHD 104 .7
Mental Illness 297 2 .0
Poor Hearing 135 .9
Stuttering 1,066 7.0
Demographics
Race
White 12,888 62.2
Nonwhite 7,842 37.8
Gender
Male 10,263 49.5
Female 10,480 50.5
Mean SD
Covariates
Low Self-control (Wave 1) 48.4 7.98
Polydrug Use (Wave 3) 1.20 1.07
Age (Wave 1) 16.2 1.74
Self-control and Health Outcomes
20
scale: all odds ratios were above 1.0, and
4 of the 5 were statistically significant at
the P < .05 level, 2-tailed tests. In this
case, because the low self-control scale
was coded such that respondents with
lower levels received higher scores, the
results indicated that lower levels of self-
control significantly increased the odds of
being diagnosed with 4 of the outcomes.
Lower levels of self-control were unre-
lated to the odds of having diabetes. Recall
that polydrug use was hypothesized as a
mediator of the relationship between self-
control and the various outcomes. As can
be seen, the polydrug use index did not
completely account for the effects of self-
control. Indeed, the polydrug use index
did not systematically reduce the effect of
self-control across the models.
To further examine the association
between levels of self-control and the odds
of developing a health-related problem,
we plotted the predicted probabilities of
being diagnosed with the 4 health-related
disorders that were significantly related
to self-control. These predicted probabili-
ties were estimated from the logistic re-
gression equations presented in Table 2.
Recall that our self-control scale was coded
so that larger values reflected lower levels
of self-control. As can be seen in Figure 1,
respondents with higher levels of self-
Table 2
Logistic Regression Models Predicting Wave 3
Health-Related Problems
Low Self-control Age Nonwhite Gender Polydrug Use
Asthma
β.009 -.058 .036 -.122 .089
SE .003 .014 .052 .049 .022
Odds Ratio 1.009** .944** 1.037 .885* 1.093**
95% C.I. 1.003 - 1.015 .918 - .970 .936 - 1.148 .804 - .975 1.046 - 1.141
Cancer
β.026 .135 -.117 -.521 .181
SE .012 .058 .213 .202 .084
Odds Ratio 1.026* 1.145* .890 .594* 1.199*
95% C.I. 1.002 - 1.051 1.022 - 1.283 .587 - 1.350 .400 - .882 1.018 - 1.412
Diabetes
β.004 .197 -.033 -.789 -.056
SE .012 .055 .195 .202 .095
Odds Ratio 1.004 1.218** .967 .454** .946
95% C.I. .981 - 1.027 1.092 - 1.358 .660 - 1.418 .306 - .674 .785 - 1.139
High Cholesterol
β.020 .065 -.019 -.079 .004
SE .006 .027 .096 .091 .043
Odds Ratio 1.020** 1.067* .981 .924 1.004
95% C.I. 1.008 - 1.031 1.013 - 1.124 .812 - 1.185 .773 - 1.104 .923 - 1.091
High Blood Pressure
β.019 .059 .221 -.050 -.068
SE .005 .024 .084 .082 .040
Odds Ratio 1.019** 1.061* 1.248** .951 .934
95% C.I. 1.009 - 1.030 1.012 - 1.112 1.058 - 1.471 .811 - 1.116 .864 - 1.011
* P<.05, 2-tailed
** P<.01, 2-tailed
Miller et al
Am J Health Behav.™™
™™
™ 2011;35(1):15-27 21
control—those scoring the lowest on the
scale—had much lower predicted prob-
abilities of having each of the 4 outcomes.
However, respondents with lower levels of
self-control—represented by higher val-
ues—had larger predicted probabilities of
having asthma, cancer, high cholesterol,
and high blood pressure. For example,
respondents with relatively high levels of
self-control had approximately a 2.2%
chance of having cancer, whereas those
with relatively lower levels of self-control
had probabilities almost twice as high
(approximately 4%).
Table 3 shows the results of the logistic
regression models that were used to pre-
dict the probabilities of having the 5 brain-
based disorders. In all 5 models, the low
self-control scale had a statistically sig-
nificant and positive effect on the odds of
having the 5 brain-based disorders. Spe-
cifically, respondents with lower levels of
self-control had higher odds of suffering
from depression, ADHD, a mental illness,
poor hearing, and stuttering speech pat-
terns. Again, all odds ratios are above 1.0
and are statistically significant at the
P<.05 level, 2-tailed tests.
Figure 2 presents a graphical repre-
sentation of the predicted probabilities of
having each of the 5 brain-based disor-
ders relative to varying levels of self-
control. As shown, lower levels of self-
control (larger values) was positively re-
0
0.05
0.1
0.15
0.2
0.25
20 25 30 35 40 45 50 55 60 65 70 75 81
Scores on the Low Self-Control Scale
Predicted Probability
Asthma
Blood Pressure
Cholesterol
Cancer
Figure 1
The Predicted Probabilities of Developing the Various Health-related
Problems at Wave III Based on Wave I Levels of Self-control
Note.
Predicted probabilities were calculated with all other variables at their mean.
Self-control and Health Outcomes
22
lated to the probability of being diagnosed
with depression. Respondents with the
highest levels of self-control—those scor-
ing the lowest on the scale—had less than
a 5% probability of being diagnosed with
depression. On the other hand, respon-
dents with the lowest levels of self-con-
trol—those scoring the highest values—
had roughly a 40% probability of being
diagnosed with depression. Although the
magnitude of probability is highest for
depression, the same exponential in-
creases are observed for the other 4 con-
ditions (ie, ADHD, mental illness, poor
hearing, and stuttering). In each case,
respondents with high self-control had a
near zero probability of having the par-
ticular brain-based disorder, whereas re-
spondents with lower levels of self-control
showed exponential increases in their
odds of having each of the brain-based
disorders.
DISCUSSION
Gottfredson and Hirschi’s general
theory of crime offers a useful framework
for understanding and predicting a wide
range of problem behaviors. Empirical
evidence has largely supported the main
propositions of the theory,2 but fewer stud-
ies have examined the relationship be-
tween low self-control and other adverse
outcomes. Those not capable of control-
ling impulses or estimating consequences
Table 3
Logistic Regression Models Predicting Wave 3 Brain-based Disorders
Low Self-control Age Nonwhite Gender Polydrug Use
Depression
β.054 -.039 -.649 -.866 .270
SE .004 .019 .076 .067 .027
Odds Ratio 1.055** .962* .523** .421** 1.309**
95% C.I. 1.047 - 1.064 .928 - .998 .450 - .607 .369 - .480 1.243 - 1.379
ADHD
β.057 -.128 -1.136 .169 .260
SE .014 .070 .359 .236 .089
Odds Ratio 1.059** .880 .321** 1.184 1.297**
95% C.I. 1.030 - 1.088 .768 - 1.009 .159 - .650 .746 - 1.880 1.090 - 1.543
Mental Illness
β.058 -.191 -.005 -.302 .243
SE .008 .042 .154 .143 .056
Odds Ratio 1.059** .826** .995 .739* 1.275**
95% C.I. 1.042 - 1.077 .761 - .897 .736 - 1.346 .558 - .979 1.143 - 1.423
Poor Hearing
β.031 .050 -.297 .447 .101
SE .012 .058 .224 .200 .084
Odds Ratio 1.032* 1.052 .743 1.563* 1.106
95% C.I. 1.007 - 1.056 .939 - 1.178 .479 - 1.153 1.057 - 2.313 .938 - 1.305
Stuttering
β.028 .018 .270 .477 .094
SE .005 .022 .076 .074 .033
Odds Ratio 1.028** 1.018 1.310** 1.611** 1.099**
95% C.I. 1.019 - 1.037 .976 - 1.062 1.128 - 1.521 1.393 - 1.864 1.031 - 1.171
* P<.05, 2-tailed
** P<.01, 2-tailed
Miller et al
Am J Health Behav.™™
™™
™ 2011;35(1):15-27 23
are more at risk for a range of deleterious
conditions including addiction, eating dis-
orders, gambling problems, and maladap-
tive sexual behaviors. Using these find-
ings as a springboard, we hypothesized
that a lack of self-control should affect
health functioning through lifestyle hab-
its and choices. The results generated
from the current study provide support for
this hypothesis and suggest that low self-
control is associated with a number of
physical and brain-based health condi-
tions.
Just as important as the health condi-
tions that were predicted by low self-con-
trol is the one that was not predicted—
diabetes. Although someone with low self-
control may have greater odds of develop-
ing high blood pressure due to a lifestyle
of impulsivity and propensity toward risk
seeking, this same person does not ap-
pear at greater risk for developing diabe-
tes. Given the prevalence of diabetes
among ethnic minorities,41 it is surpris-
ing that minority respondents were no
more likely to have been diagnosed. In
fact, minority respondents were no more
likely to have been diagnosed with any of
the health-related problems with the ex-
ception of high blood pressure. Secondary
analyses not reported here indicated that
only African Americans and Native Ameri-
cans had greater odds of being diagnosed
with high blood pressure. This particular
finding is consistent with the Center of
Disease Control national-level data as
well, which finds that African Americans
and Native Americans are significantly
Figure 2
The Predicted Probabilities of Developing the Various Health-related
Problems at Wave III Based on Wave I Levels of Self-control
Note.
Predicted probabilities were calculated with all other variables at their mean.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
20 25 30 35 40 45 50 55 60 65 70 75 81
Scores on the Low Self-Control Scale
Predicted Probabilit
y
Depression
Stuttering
Mental Illness
Poor Hearing
ADHD
Self-control and Health Outcomes
24
more likely to suffer from high blood pres-
sure.42 Similarly, secondary analyses in-
dicated that Native Americans had greater
odds of suffering from stuttering speech
patterns.
Overall, findings from this study sup-
ported our hypothesis that self-control is
associated with a wide range of both physi-
cal and brain-based disorders. Several
theoretical conclusions can be drawn from
these analyses. First, lower levels of self-
control were related to nearly all of the
adverse health conditions, suggesting
that the trait is consistently associated
with negative outcomes as they relate to
dimensions of physical and psychological
well-being. This offers empirical evidence
in support of Gottfredson and Hirschi’s
prediction that persons with relatively
low levels of self-control are far more
likely to suffer from consequences be-
yond those of crime and delinquency, in
this case, adverse health. Gottfredson
and Hirschi speculated:
…the ‘costs’ of low self-control for the
individual may far exceed the costs of
his criminal acts. In fact, it appears
that crime is often among the least
serious consequences of a lack of self-
control in terms of the quality of life of
those lacking it.12
Second, these findings reveal that those
with low self-control are exponentially more
likely to suffer from most of the conditions
examined (again, with the exception of
diabetes), evidencing the significance of
the trait relative to other variables. It is
likely that low self-control operates
through situational and contextual fac-
tors that place those individuals at greater
risk of, or opportunity for, behavior that is
ultimately deleterious to one’s health.32
Ultimately, the culmination of a lifetime,
or in the current study, an adolescence
and early adulthood, fraught with impul-
sivity, risk taking, and poor decision
making leads to the myriad negative con-
sequences hypothesized by Gottfredson
and Hirschi.12
Though this study offers an initial ex-
ploration into the relationship between
self-control and health outcomes, self-
control was not found to fully mediate the
effects of the other independent vari-
ables. This suggests that many of the
known correlates of health, such as race
and gender, still play a significant role in
the etiology of adverse health, net of the
effects of self-control. Also, the polydrug
use index exhibited a significant effect on
many of the health outcomes but was
unable to fully account for the effects of
self-control. This suggests that other
factors and life choices may also mediate
the relationship between self-control and
adverse health outcomes. For example,
levels of self-control differentiated respon-
dents with and without hearing loss. Al-
though problems with hearing can mani-
fest due to factors unrelated to one’s level
of self-control, damage to hearing can
also occur if a person does not take proper
precaution in certain environments (eg,
wearing earplugs in noisy environments).
Perhaps this begins to explain the rela-
tionship observed in the current study.
The lack of foresight exercised by indi-
viduals with lower levels of self-control, as
well as the heightened tendency toward
risk seeking that is more prevalent in
those with low self-control, might increase
the likelihood of exposure to harmful en-
vironments (eg, noisy environments),
which in turn increases the odds of suf-
fering from hearing loss. Future research-
ers may wish to build on this work by
identifying these mediating factors. It is
also possible that self-control interacts
with demographics and other contextual
factors to better explain the link between
self-control and health. Because this
study is exploratory in nature, we refrain
from specifically testing some of these
more complex and perhaps interactive
relationships.
We present the current findings as an
initial step toward understanding the
ways in which levels of self-control im-
pinge upon everyday life. As alluded to
above, it is not at all clear that levels of
self-control directly affect any of the ob-
served outcomes. Instead, it is more likely
that a person’s level of self-control is
merely one exogenous factor that in-
creases the odds of exposure to risky
environments that increase the chances
of health problems. It may also be the case
that levels of self-control are related to
some health or brain-based disorders due
to shared etiological pathways. For ex-
ample, the current study found evidence
of a link between levels of self-control and
problems with stuttering. Recent research
has shown that the genetic factors affect-
ing levels of self-control overlap with those
affecting language deficits.43 Thus, it is
Miller et al
Am J Health Behav.™™
™™
™ 2011;35(1):15-27 25
possible that the association between stut-
tering and levels of self-control may be
mediated genetically, environmentally,
or by some combination of the 2; a possi-
bility that holds for all of the other out-
comes as well. Similarly, this analysis
revealed a link between low self-control
and depression. Though depression is
believed to be caused by several factors,
some of which may interact, one frame-
work views it as the result of an inability
to produce positive experiences and to
avoid aversive ones.44 These are 2 traits
that are consistent with low self-control.
Additionally, some view depression as the
result of interpersonal discord,45 which is
also predicted by Gottfredson and Hirschi’s
theory.12
Strategies aimed at targeting the de-
velopment of self-control during child-
hood and adolescence may aid in the
prevention of future health problems.
Bandura argues that preventative efforts
aimed at health functioning are espe-
cially important because patterns of be-
havior that can compromise health typi-
cally begin in early adulthood and con-
tinue throughout the life course.32(p301) By
equipping children and adolescents with
skills that enable them to regulate their
emotional states and manage diverse
stressors, they are not only more likely to
exert control over the quality of their
health, but also benefit more holistically
from enhanced self-control. Low self-con-
trol is not merely associated with nega-
tive outcomes; the reverse is also true.
Children and adolescents with high lev-
els of self-control are more likely to have
higher grade point averages, better ad-
justment, higher self-esteem, less psy-
chopathology, less binge eating and alco-
hol abuse, better relationships and inter-
personal skills, and more secure attach-
ments.30 Thus, improving self-control and
self-regulation may produce an exponen-
tial positive impact across a constellation
of life dimensions.
REFERENCES
1.Gottfredson M. The empirical status of control
theory in criminology. In Cullen, FT, Wright,
JP, Blevins, KR, (Eds). Taking Stock: The
Status of Criminological Theory, Vol. 15.
Edison, NJ: Transaction Publishers 2006:77-
100.
2.Pratt TC, Cullen FT. The empirical status of
Gottfredson and Hirschi’s general theory of
crime: a meta-analysis. Criminology.
2000;38:931-964.
3.Evans DT, Cullen FT, Burton VS Jr., et al. The
social consequences of self-control: testing
the General Theory of Crime. Criminology.
1997;35:475-504.
4.Junger M, Tremblay RE. Self-control, acci-
dents, and crime. Criminal Justice and Behav-
ior. 1999;26:485-501.
5.Junger M, van der Heijden P, Keane C. Inter-
related harms: examining the associations
between victimization, accidents, and crimi-
nal behavior. Injury Control and Safety Promo-
tion. 2001;8:13-28.
6.Jessor R, Jessor SJ. Problem Behaviour and
Psychosocial Development. New York: Aca-
demic Press 1977.
7.Osgood DW, Johnston LD, O’Malley PM, et al.
The generality of deviance in late adolescence
and early adulthood. American Sociological Re-
view. 1988;53:81-93.
8.Donovan JE, Jessor R. Structure of problem
behavior in adolescence and young adult-
hood. J Consult Clin Psychol. 1985;53:890-
904.
9.Donovan, J, Jessor R, Costa FM. Adolescent
health behavior and conventionality-uncon-
ventionality: an extension of problem-behav-
ior theory. Health Psychol. 1988;10:52-61.
10.Robins LN, Wish E. Childhood deviance as a
developmental process. A study of 233 urban
black male youth from birth to 18. Social
Forces. 1977;34:448-4473.
11.Farrington DP. Crime and physical health:
illness, injuries, accidents, and offending in
the Cambridge study. Crim Behav Ment Health.
1995;5:261-278.
12.Gottfredson M, Hirschi T. A General Theory
of Crime. Palo Alto, CA: Stanford University
Press 1990.
13.Piquero A, Tibbetts S. Specifying the direct
and indirect effects of low self-control and
situational factors in decision-making: to-
ward a more complete model of rational of-
fending. Justice Quarterly. 1996;13:481-510.
14.Paternoster R, Brame R. The structural simi-
larity of processes generating criminal and
analogous behaviors. Criminology. 1998;36:633-
670.
15.Shepherd J, Farrington DP, Potts J. Impact of
antisocial lifestyle on health. J Public Health.
2004;26:347-352.
16.Laub JH, Vaillant GE. Delinquency and mor-
tality: a 50-year follow-up study of 1,000
delinquent and non-delinquent boys. Am J
Psychol. 2000;157:96-102.
17.Rutter M, Giller H. Juvenile Delinquency,
Trends and Perspectives. Harmondsworth:
Penguin Books 1983.
18.Werner EE. Vulnerability and resiliency in
children at risk for delinquency: a longitudi-
nal study from birth to adulthood. In Burchard
JD, Burchard SN, (Eds). Prevention of Delin-
quent Behavior. Newbury Park: Sage 1987:16-
43.
19.Wilson JQ, Herrnstein RJ. Crime and Human
Nature. New York: Simon & Schuster 1985.
Self-control and Health Outcomes
26
20.Junger M, Stroebe W, van der Laan AM.
Delinquency, health behavior and health. Brit
J Health Psychol. 2001;6:103-120.
21.Samuelson YM, Hodgins S, Larsson A, et al.
Adolescent antisocial behavior as predictor of
adverse outcomes to age 50. Criminal Justice
and Behavior. 2009;37(1):1-17.
22.Connor BT, Stein JA, Longshore D. Examin-
ing self-control as a multidimensional predic-
tor of crime and drug use in adolescents with
criminal histories. J Behav Health Serv Res.
2009;36(2):137-149.
23.Suchman EA. Accidents and social deviance.
J Health Soc Behav. 1970;11:4-15.
24.Baumeister RF, Heatherton TF. Self-regula-
tion failure: an overview. Psychological Inquiry.
1996;7:1-15.
25.Mischel W, Shoda Y, Peake P. The nature of
adolescent competencies predicted by pre-
school delay of gratification. J Pers Soc Psychol.
1988;54:687-696.
26.Baumeister RF, Bratlavsky E, Muraven M, et
al. Ego depletion: is the active self a limited
resource? In Baumeister RF, (Ed). The Self in
Social Psychology. Philadelphia, PA: Psy-
chology Press 1998:317-336.
27.Becker HM. The Health Belief Model and
Personal Health Behavior. Thorofare, NJ:
Slack 1974.
28.Godin G, Kok GJ. The theory of planned
behavior: a review of its application to health-
related behaviors. Am J Health
Promot.1996;11:87-98.
29.Peluso T, Ricciardelli LA, Williams RJ. Self-
control in relation to problem drinking and
symptoms of disordered eating. Addict Behav.
1999;24:439-442.
30.Tangney JP, Baumeister RF, Boone AL. High
self-control predicts good adjustment, less
pathology, better grades, and interpersonal
success. J Pers. 2004;72:271-322.
31.Schroeder KEE, Schwarzer R. Habitual self-
control and the management of health behav-
ior among heart patients. Soc Sci Med.
2005;60:859-875.
32.Bandura A. Self-Efficacy: The Exercise of
Control. New York: W.H. Freeman and Com-
pany 1997.
33.Schroeder KEE. Self-regulation Competence
in Coping with Chronic Disease. New York:
Waxmann Publishing 1997.
34.Udry JR. The National Longitudinal Study of
Adolescent Health (Add Health), Waves I and
II, 1994–1996; Wave III, 2001–2002 [ma-
chine-readable data file and documentation].
Chapel Hill, NC: Carolina Population Center,
University of North Carolina at Chapel Hill
2003.
35.Harris KM, Florey F, Tabor J, et al. The
National Longitudinal Study of Adolescent
Health: Research Design [online]. Available
at: http://www.cpc.unc.edu/projects/
addhealth/design. Accessed July 15, 2009.
36.DeLisi M, Hochstetler A, Murphy DS. Self-
control behind bars: a validation study of the
Grasmick et al. scale. Justice Quarterly.
2003;20:241-263.
37.Longshore D, Turner S, Stein JA. Self-control
in a criminal sample: an examination of
construct validity. Criminology. 1996;34:209-
228.
38.Marcus B. An empirical examination of the
construct validity of two alternative self-con-
trol measures. Education and Psychological
Measurement. 2003;63:674-706.
39.Grasmick HG, Tittle CR, Bursik RJ, et al.
Testing the core empirical implications of
Gottfredson and Hirschi’s general theory of
crime. Journal of Research in Crime and Delin-
quency. 1993;30:5-29.
40.Perrone D, Sullivan CJ, Pratt TC, et al.
Parental efficacy, self-control, and delin-
quency: a test of a General Theory of Crime on
a nationally representative sample of youth.
Int J Offender Ther Comp Criminol. 2004;48:298-
312.
41.Centers for Disease Control and Prevention.
National diabetes fact sheet: General informa-
tion and national estimates on diabetes in the
United States, 2007. Atlanta, GA, U.S. De-
partment of Health and Human Services,
Centers for Disease Control and Prevention,
2008.
42.National Center for Health Statistics. Health,
United States, 2005 (on-line). Available at:
http://www.cdc.gov/nchs/hus.htm. 2006.
Accessed July 20, 2009.
43.Beaver KM, DeLisi M, Vaughn MG, et al. The
relationship between self-control and lan-
guage: evidence of a shared etiological path-
way. Criminology. 2008;46:201-223.
44.Lewinsohn PM, Hoberman HM, Hautzinger
M. An integrated theory of depression. In
Reiss S, Bootzin R, (Eds). Theoretical Issues
in Behavior Therapy. New York: Academic
Press 1985:331-359.
45.Coyne JC. Interpersonal processes in depres-
sion. In Keitner GI, (Ed). Depression and
Families: Impact and Treatment. Washington,
DC: American Psychiatric Press 1990:33-55.
Miller et al
Am J Health Behav.™™
™™
™ 2011;35(1):15-27 27
Appendix A
Description of Variables and Scales
Low Self-control Scale (Wave I)
Scale created by summing the following self-reported measures (*=question was asked to the parent):
1. All things considered, how is your child’s life going?* (0=very well; 1=fairly well; 2=not so well; 3=not well at all)
2. Do you get along well with your child?* (0=always; 1=often; 2=sometimes; 3=seldom; 4=never)
3. Do you feel you can trust your child?* (0=always; 1=often; 2=sometimes; 3=seldom; 4=never)
4. Does your child have a bad temper?* (0=no; 1=yes)
5. You never argue with anyone? (0=strongly agree; 1=agree; 2=neutral; 3=disagree; 4=strongly disagree)
6. When you get what you want, it’s usually because you worked hard for it? (0=strongly agree; 1=agree; 2=neutral; 3=disagree;
4=strongly disagree)
7. You never get sad? (0=strongly agree; 1=agree; 2=neutral; 3=disagree; 4=strongly disagree)
8. You never criticize other people? (0=strongly agree; 1=agree; 2=neutral; 3=disagree; 4=strongly disagree)
9. You usually go out of your way to avoid having to deal with problems in your life? (0=strongly agree; 1=agree; 2=neutral; 3=disagree;
4=strongly disagree)
10. Difficult problems make you very upset? (0=strongly disagree; 1=disagree; 2=neutral; 3=agree; 4=strongly agree)
11. When making decisions, you usually go with your “gut feeling” without thinking too much about the consequences of each alternative?
(0=strongly disagree; 1=disagree; 2=neutral; 3=agree; 4=strongly agree)
12. When you have a problem to solve, one of the first things you do is get as many facts about the problem as possible? (0=strongly agree;
1=agree; 2=neutral; 3=disagree; 4=strongly disagree)
13. When attempting to find a solution to a problem, you usually try to think of as many different ways to approach the problem as possible?
(0=strongly agree; 1=agree; 2=neutral; 3=disagree; 4=strongly disagree)
14. When making decisions, you generally use a systematic method for judging and comparing alternatives? (0=strongly agree; 1=agree;
2=neutral; 3=disagree; 4=strongly disagree)
15. After carrying out a solution to a problem, you usually try to analyze what went right and what went wrong? (0=strongly agree; 1=agree;
2=neutral; 3=disagree; 4=strongly disagree)
16. You like yourself just the way you are? (0=strongly agree; 1=agree; 2=neutral; 3=disagree; 4=strongly disagree)
17. You feel like you are doing everything just about right? (0=strongly agree; 1=agree; 2=neutral; 3=disagree; 4=strongly disagree)
18. You feel socially accepted? (0=strongly agree; 1=agree; 2=neutral; 3=disagree; 4=strongly disagree)
19. Do you have trouble getting along with your teachers? (0=never; 1=just a few times; 2=about once a week; 3=almost everyday;
4=everyday)
20. Do you have trouble paying attention in school? (0=never; 1=just a few times; 2=about once a week; 3=almost everyday; 4=everyday)
21. During the past week, did you have trouble keeping your mind on what you were doing? (0=never; 1=sometimes; 2=a lot of the time;
3=most or all of the time)
22. Do you have trouble getting your homework done? (0=never; 1=just a few times; 2=about once a week; 3=almost everyday; 4=everyday)
23. Do you have trouble getting along with other students? (0=never; 1=just a few times; 2=about once a week; 3=almost everyday;
4=everyday)
Polydrug Use Index (Wave III)
Scale created by summing the following self-reported measures (0=no; 1=yes):
1. In the past year, have you used marijuana?
2. In the past year, have you used any kind of cocaine?
3. In the past year, have you used crystal meth?
4. In the past year, have you used any of these types of illegal drugs (LSD, PCP, ecstasy, mushrooms, inhalants, ice, heroin, or prescription
medicines not prescribed to you)?
5. In the past year, have you injected an illegal drug?
6. During the past 12 months, on how many days did you drink alcohol? (0=none; 1=at least once)
Health-Related Problems – Individual Items (Wave III)
Responses coded 0=no; 1=yes:
1. Asthma – Have you ever been diagnosed with asthma?
2. Cancer – Have you ever been diagnosed with cancer or leukemia?
3. Diabetes – Have you ever been diagnosed with diabetes?
4. High Cholesterol – Has a doctor ever told you that you have high cholesterol?
5. High Blood Pressure – Have you ever been diagnosed with high blood pressure?
Brain-Based Disorders – Individual Items (Wave III)
Responses coded 0=no; 1=yes:
1. Depression – Have you ever been diagnosed with depression?
2. ADHD – Have you taken prescription medication in the past 12 months for ADD or ADHD?
3. Mental Illness – In the past 5 years, have you spent a day or more in a facility where you were treated for a mental illness?
4. Poor Hearing – How is your hearing? If you wear a hearing aid, describe your hearing without it. (0=excellent, good, or fair; 1=poor,
very poor, or deaf)
5. Stuttering – Do you have a problem with stuttering or stammering?
