PDF Archive

Easily share your PDF documents with your contacts, on the Web and Social Networks.

Send a file File manager PDF Toolbox Search Help Contact



Genetic Risk African American .pdf



Original filename: Genetic Risk African American.pdf
Title:
Author: Kevin M. Beaver; Ashley Sak; Jamie Vaske; Jessica Nilsson

This PDF 1.4 document has been generated by Elsevier / Unknown, and has been sent on pdf-archive.com on 11/05/2011 at 02:11, from IP address 81.170.x.x. The current document download page has been viewed 1289 times.
File size: 165 KB (5 pages).
Privacy: public file




Download original PDF file









Document preview


Psychiatry Research 175 (2010) 160–164

Contents lists available at ScienceDirect

Psychiatry Research
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p s yc h r e s

Genetic risk, parent–child relations, and antisocial phenotypes in a sample of
African-American males
Kevin M. Beaver a,⁎, Ashley Sak a, Jamie Vaske b, Jessica Nilsson a
a
b

College of Criminology and Criminal Justice, Florida State University, Tallahassee, FL 32306, United States
Division of Criminal Justice, University of Cincinnati, Cincinnati, OH 45221-0389, United States

a r t i c l e

i n f o

Article history:
Received 10 August 2008
Received in revised form 15 January 2009
Accepted 22 January 2009
Keywords:
Antisocial behavior
Genetics
Environment
Interactions

a b s t r a c t
Gene × environment interactions have been found to be associated with the development of antisocial
behaviors. The extant gene × environment research, however, has failed to measure directly the ways in
which global measures of genetic risk may interact with a putative environmental risk factor. The current
study addresses this gap in the literature and examines the interrelationships among a global measure of
genetic risk based on five genetic polymorphisms, a measure of parent–child relations, and eight antisocial
phenotypes. Analysis of African-American males (N = 145 to 159) drawn from the National Longitudinal
Study of Adolescent Health (Add Health) revealed two broad findings. First, the genetic risk and parent–child
relations scales were inconsistently related to the outcome variables. Second, genetic risk and parent–child
relations interacted to predict variation in all of the eight antisocial phenotype measures. These findings
point to the possibility that measures of genetic risk that are based on multiple polymorphisms can be
employed to examine the gene × environmental basis to antisocial behavioral phenotypes.
© 2009 Elsevier Ireland Ltd. All rights reserved.

1. Introduction
Antisocial behaviors, including crime and delinquency, are the result
of a complex arrangement of genetic and environmental factors working
independently and synergistically (Moffitt, 2005; Rutter, 2006). Results
from twin and adoption studies, for example, have revealed that between
40 to 80% of the variance in antisocial phenotypes is attributable to
genetic factors, with most of the remaining variance accounted for by the
nonshared environment (Mason and Frick, 1994; Miles and Carey, 1997;
Rhee and Waldman, 2002; Arseneault et al., 2003). But partitioning
variance into genetic and environmental components fails to capture the
ways in which the environment moderates genetic effects and the ways
in which genetic factors moderate environmental influences. There is,
however, an ever-expanding line of molecular genetic research that has
examined gene× environment interactions in the etiology of antisocial
phenotypes (Caspi et al., 2002; Kim-Cohen et al., 2006). This line of
research has provided relatively convincing evidence that certain genes
interact with certain environments to produce an array of criminal and
delinquent behaviors.
For instance, in a landmark study, Caspi et al. (2002) examined the
effect that a polymorphism in the promoter region of the MAOA gene
had on antisocial phenotypes. This research team hypothesized that the
low-activity MAOA allele would only have an effect for those subjects
who had a history of childhood maltreatment. The results of the study

⁎ Corresponding author. Tel.: +1 850 644 9180; fax: +1 850 644 9614.
E-mail address: kbeaver@fsu.edu (K.M. Beaver).
0165-1781/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.psychres.2009.01.024

supported their hypothesis. MAOA was not associated with antisocial
behaviors for those males who had not been maltreated. A very different
set of findings emerged for males who had been maltreated. Although
only 12% of the sample had been maltreated and possessed the lowactivity MAOA allele, they accounted for 44% of all criminal convictions,
and 85% of those subjects who had been severely maltreated and who
had the low-activity MAOA allele displayed antisocial behavior. Followup studies have both replicated (Foley et al., 2004; Kim-Cohen et al.,
2006; Widom and Brzustowicz, 2006; Frazzetto et al., 2007) and failed
to replicate (Haberstick et al., 2005; Huizinga et al., 2006; Young et al.,
2006) this gene× environment interaction. Other genes—especially
those of the dopaminergic and serotonergic systems—have also been
found to interact with certain environmental stimuli in the prediction of
a wide range of antisocial phenotypes (Nilsson et al., 2005; BakermansKranenburg and van IJzendoorn, 2006).
The gene × environment literature has provided a great deal of
insight about which particular genes interact with which particular
environments to produce different behaviors. There is, however, at
least one main limitation with the gene × environment research.
Antisocial behaviors are polygenic phenotypes, but the existing
gene × environment literature fails to model polygenic effects. Instead,
most extant research artificially divorces a particular gene from the
larger genome and examines how this isolated genetic factor interacts
with the environment. While this approach is useful in determining
which genes are associated with antisocial phenotypes, it cannot
measure each subject's total genetic risk.
The present study addresses this gap in the gene × environment
research and develops a measure of genetic risk that is based on five

K.M. Beaver et al. / Psychiatry Research 175 (2010) 160–164

polymorphisms. We then examine whether this measure of genetic
risk has an effect on eight different antisocial phenotypes. Research
has revealed that adverse home environments interact with genetic
factors to predict antisocial behavior (Cadoret et al., 2003; Tuvblad
et al., 2006). As a result, we also examine whether a measure of
parent–child relations moderates the effects of genetic risk. Drawing
from the existing literature, we hypothesize that the genetic risk
measure will only be related to antisocial phenotypes among those
subjects who had negative parent–child relations.
2. Methods
2.1. Study participants
Subjects for this study came from the National Longitudinal Study of Adolescent
Health (Add Health) (Udry, 2003). The Add Health is a longitudinal, nationallyrepresentative, and prospective sample of American youths who were enrolled in
seventh through twelfth grade in 1994–1995. Overall, the subjects have been assessed
at three different time periods. The first assessments were gathered when the subjects
were between the ages of 11 and 19 years old, where a total of more than 90,000
students were interviewed at school. A stratified subsample of more than 20,700 youths
was then selected to participate in the wave 1 in-home survey during that same school
year. The subject's primary caregiver was also interviewed and provided detailed
information about the respondent. Approximately 1 to 2 years later, during the second
wave of interviews, 14,738 of these subjects were reassessed. The third round of
assessments was conducted in 2001–2002 when most of the respondents were
between the ages of 18 to 27 years old. During this last assessment period, 15,197 of the
original respondents were re-interviewed. More detailed information about the
sampling design has been published elsewhere (Resnick et al, 1997; Harris et al., 2003).
Sibling pairs were oversampled in the Add Health study. During the initial
assessments, subjects were asked whether they currently resided with a co-twin, a halfsibling, a cousin, or a step-sibling. If they responded affirmatively, then their sibling was
recruited to participate in the study. A probability sample of full siblings was also added
to the study. Overall, 5470 siblings were included in the Add Health sample. Then,
during the last wave of assessments, subjects who were part of the sibling pairs
subsample were asked to provide samples of their DNA for genotyping. In total, 2574
respondents were included in the DNA subsample of the Add Health study (Haberstick
et al., 2005).
The sample for the current study was confined to African-American males. Three
reasons informed the decision to focus on this group. First, for females, the base-rate of
offending was extremely low in the Add Health sample, especially in adulthood. As
such, females were excluded from the analysis. Second, population stratification effects
can result when subjects from different racial groups are analyzed simultaneously. To
control for this potential problem, all of the models were calculated separately for nonHispanic Caucasians and for African Americans. The results of these analyses did not
reveal any statistically significant effects for Caucasian males. Third, there has been a
lack of research examining the genetic contributors to antisocial behavior in samples of
African-American males. The current study addresses this gap in the literature by
analyzing data drawn solely from African-American males.

161

MAOA was genotyped for a 30 base pair VNTR in the promoter region of the gene.
Allele lengths ranged from 2- to 5-repeats in the sample. Using the same coding scheme
as Haberstick et al. (2005), these alleles were then divided into two groups. The first
group, which was scored as the group of risk alleles, contained the 2-repeat allele and
the 3-repeat allele. The second group contained the 3.5-, 4-, and 5-repeat alleles. All five
of the polymorphisms were in Hardy–Weinberg equilibrium.
2.3. Genetic risk
Prior researchers have developed innovative methods for measuring overall
assessments of an individual's genetic risk indirectly (Kendler et al., 1995; Jaffee et
al., 2005), but not directly. It is possible, however, to create a composite measure of
genetic risk by using the genetic polymorphisms described above. All of these
polymorphisms have been linked to different types of antisocial behaviors, including
physical violence (Beaver et al., 2007; Guo et al., 2007a) and alcohol abuse (Guo et al.,
2007b; Hopfer et al., 2005) among subjects from the Add Health study. To create the
measure of genetic risk, the five polymorphisms were coded co-dominantly, where the
score for each genetic variable measured the number of risk alleles that were present
for that particular gene. Then, all of the polymorphisms were summed together, which
resulted in a genetic risk scale that indexed the total number of risk alleles that each
subject possessed. Scores on the genetic risk scale ranged between 0 and 9, with higher
scores reflecting more genetic risk for antisocial behavior.
2.4. Assessment of parent–child relations
Three different parent–child relationship scales, all of which were measured at
wave 1, were available in the Add Health data. The first scale, maternal involvement,
indexes the extent to which the mother is involved in the subject's life. During wave 1
interviews, subjects were presented with a list of 10 different activities and asked
which, if any, they had done with their mother in the past month. Respondents, for
example, were asked whether they had gone shopping with their mother, whether they
had gone to a movie with their mother, and whether they had played a sport with their
mother. Responses to each question were coded dichotomously (0 = yes, 1 = no). The
items were then summed together, where higher scores indicate less maternal
involvement (Cronbach's α = 0.48).
The second scale, maternal attachment, is a two-item measure that captures how
strongly bonded the subject is with their mother. At wave 1, subjects were asked how
close they felt with their mother and how much they thought that their mother cared
about them. These two items were then summed together and reverse-coded, such that
higher scores indicate less maternal attachment (Cronbach's α = 0.77). This same scale
has been used by previous researchers (Haynie, 2001; Schreck et al., 2004).
Maternal disengagement is the third and final parent–child relations scale. This
five-item measure indexes the degree to which the subject's parents were disengaged
from them. During wave 1 interviews, subjects were asked to indicate how warm and
loving their mother was, how often they talked with their mother, and the overall
quality of their relationship, among others. These five items were then added together,
where higher scores reflect more maternal disengagement (Cronbach's α = 0.78).
These three scales were then subjected to a principal components factor analysis with
varimax rotation. The results revealed that all of the scales loaded on a unitary factor
and, consequently, the scales were transformed into a weighted factor score that
provides a measure of parent–child relations.
2.5. Assessment of antisocial phenotypes

2.2. Genotyping
Subjects submitted buccal cells to be genotyped for polymorphisms found in the
following five genes: the dopamine D2 receptor gene (DRD2), the dopamine D4
receptor gene (DRD4), the dopamine transporter gene (DAT1), the serotonin
transporter gene (5HTT), and the monoamine oxidase A gene (MAOA). Genotyping
was performed by geneticists at the Institute for Behavioral Genetics at the University of
Colorado. Information about the genotyping protocol, including primer sequences and
amplification processes, has been described in detail in previously published outlets
(Haberstick et al., 2005; Hopfer et al., 2005; Beaver et al., 2007; Guo et al., 2007a,b).
Briefly, DRD2 was genotyped for the TaqIA polymorphism, which is located in the 3'
untranslated region of the gene. This polymorphism results from two different alleles:
the A-1 allele and the A-2 allele. The A-1 allele was scored as the risk allele.
DRD4 has a 48 base pair VNTR that can be repeated between 2 and 11 times, but the
2-, 4-, and 7-repeat alleles are the most common. Consistent with past research (Beaver
et al., 2007), the 2-, 3-, 4-, 5-, and 6-repeat alleles were pooled together into one group
and the 7-, 8-, 9-, and 10-repeat alleles were pooled together into another group. This
latter group of alleles was identified as the risk allele group.
The DAT1 polymorphism consists of a 40 base pair VNTR. Subjects were genotyped
for this polymorphism, which produced repeat sequences that ranged from six repeats
to eleven repeats. In line with prior research (Hopfer et al., 2005), subjects who
possessed alleles other than the 9- or 10-repeat allele were removed from the analysis.
This resulted in losing less than 5% of the sample. The 10-repeat allele was scored as the
risk allele.
Subjects were genotyped for the 44 base pair VNTR in the serotonin transporter
gene, which resulted in two alleles: a short allele (484 base pairs) and a long allele (528
base pairs). The short allele was identified as the risk allele.

The Add Health data contain a number of items that measure involvement in antisocial
behaviors at each round of interviews. As a result, a serious delinquency scale and a violent
delinquency scale were created at each of the three assessment periods. The wave 1 and
wave 2 serious delinquency scales contain the same eleven self-report items. Subjects, for
example, were asked to indicate how frequently in the past 12 months they damaged
property, broke into a house, sold drugs, shot or stabbed someone, and pulled knife or gun
on someone, among others. Responses to these items were then summed together to form
the wave 1 (Cronbach's α = 0.81) and wave 2 (Cronbach's α = 0.83) serious delinquency
scales. Seven of these items measured involvement in acts of serious, physical violence.
These seven items were then included in a separate violent delinquency scale at both wave
1 (Cronbach's α = 0.76) and wave 2 (Cronbach's α = 0.77). These four delinquency scales

Table 1
Descriptive statistics for the Add Health dependent variables.

Wave 1 serious delinquency
Wave 1 violent delinquency
Wave 2 serious delinquency
Wave 2 violent delinquency
Wave 3 serious delinquency
Wave 3 violent delinquency
Number of police contacts
Antisocial behavior index

Mean

Standard Deviation

Skewness

2.52
1.87
1.66
1.20
1.50
1.11
0.53
0.13

4.06
3.05
3.60
2.75
3.68
3.47
1.04
0.38

3.00
2.94
3.37
3.73
5.73
6.69
2.12
3.16

162

K.M. Beaver et al. / Psychiatry Research 175 (2010) 160–164

Table 2
The effects of genetic risk, parent–child relations, and gene × environment interactions on the wave 1 delinquency scales.
Wave 1 Serious delinquency scale
b
Genetic risk
Parent–child relations
Genetic risk × parent–child relations
Age
n

SE

Wave 1 Violent delinquency scale

P

b

SE

P

b

P

b

SE

P

0.03
0.21
0.33
0.08

0.09
0.10
0.09
0.07
153

0.784
0.039
0.000
0.253

0.11
0.17

0.09
0.12

0.243
0.174

0.06

0.08
156

0.429

0.04
0.18
0.24
0.08

0.10
0.10
0.09
0.08
156

0.713
0.070
0.008
0.340

0.09
0.19

0.09
0.14

0.277
0.178

0.06

0.07
153

0.860

SE

Notes:
Negative binomial regression equations;
Huber/White standard errors estimated.

are virtually identical to the ones constructed by Guo et al. (2007a). At wave 3, however,
not all of the same delinquency questions were asked and thus the wave 3 scales contained
a slightly different set of items that still measured serious and violent delinquency. For
example, subjects were asked how frequently they had engaged in group fights, how often
they had broken into a house, and how often they had injured someone badly enough to
need bandages. In total, twelve items were included in the wave 3 serious delinquency
scale (Cronbach's α = 0.76) and eight items were included in the wave 3 violent
delinquency scale (Cronbach's α = 0.71).
The last two antisocial phenotype variables measured lifetime involvement in
criminal and aggressive behaviors. First, at wave 3, subjects were asked to indicate the
total number of times they had been stopped by the police and questioned for reasons
other than minor traffic violations. Second, a lifetime antisocial behavior index was
created by first identifying items from the three violent delinquency scales that
approximated DSM-IV criteria for conduct disorder. These items were then summed
together and the resulting scale was transformed into a standardized score (z-score).
Subjects who had a score of 1.5 standard deviations or higher on this scale were
assigned a value of “1” and all other subjects were assigned a value of “0.” Next, all
respondents who had indicated that they had been convicted of a felony were assigned
a value of “1” and all subjects who indicated they had never been convicted of a felony
were assigned a value of “0.” Scores on these two variables (i.e., the transformed
conduct disorder scale and the felony conviction variable) were added together to form
the composite index of antisocial behavior. This scale is identical to the one used by
Beaver et al. (2007). Table 1 contains the means, standard deviations, and skewness
statistics for all eight of the antisocial phenotype measures. It should be noted, however,
that since most of the antisocial phenotype measures are highly intercorrelated they
may be tapping an underlying construct and thus are not necessarily measuring eight
distinct phenotypes.
2.6. Statistical analysis
A series of multivariate regression equations were estimated to examine the
genetic, environmental, and gene × environment effects on the eight different antisocial
phenotype measures. For each outcome measure, two equations were calculated. First,
the genetic risk scale and the parent–child relations scale were entered into the
equation. These models examined the independent effects that genes and the
environment had on antisocial behavior. Second, a gene × environment interaction
term was introduced into the equations. This interaction term was created by
multiplying the genetic risk scale and the parent–child relations scale together.
Collinearity diagnostics were calculated and the results indicated that collinearity was
not a problem in any of the models. All of the equations controlled for the subject's age.
The eight antisocial phenotype scales are count measures and thus are not normally
distributed, violating one of the main assumptions needed to employ ordinary least
squares (OLS) regression analyses. As Table 1 shows, the skewness statistics for all of
the outcome scales are well above the conventional threshold of 2.0, indicating that
they are severely skewed (Ritchey, 2000). Visual inspection of the variables indicated
that they approximate a Poisson distribution and are overdispersed. To take this into
account, all of the models were estimated using negative binomial regression.

Additionally, not all of the observations in the Add Health sample were independent
of each other (i.e., more than one sibling per household), which can result in
downwardly biased standard errors. This problem was corrected in two ways. First, one
twin from each monozygotic twin pair was randomly removed from the analysis
(Haberstick et al., 2005). Second, all of the models were estimated using Huber/White
standard errors which corrects for the clustering of observations. All test statistics are
considered statistically significant if P ≤ 0.05 (two-tailed).

3. Results
The analysis begins by examining the effects that genetic risk,
parent–child relations, and the genetic risk × parent–child relations
interaction term have on wave 1 serious delinquency and wave 1 violent
delinquency. The results of these models are displayed in Table 2. Three
findings are immediately obvious across all of the equations. First, the
genetic risk measure fails to reach statistical significance. Second, there is
not a direct association between parent–child relations and delinquency.
Third, and perhaps most important, the genetic risk× parent–child
relations interaction term is the strongest predictor of the two
delinquency scales. In substantive terms, this interaction can be
interpreted to mean that the effects of genetic risk on delinquency at
wave 1 are moderated by parent–child relations.
Table 3 presents the results of the models predicting the wave 2
serious delinquency scale and the wave 2 violent delinquency scale.
The pattern of results is strikingly similar to those reported in Table 2.
For example, genetic risk fails to have a statistically significant main
effect on either of the two delinquency scales. The parent–child
relations scale also fails to reach statistical significance in any of the
models. The genetic risk × parent–child relations interaction term,
however, approaches statistical significance (P = 0.067) in the
equation predicting the wave 2 serious delinquency scale and has a
statistically significant effect on the wave 2 violent delinquency scale.
The next set of models, which are presented in Table 4, examine
the predictors of the wave 3 serious delinquency scale and the
predictors of the wave 3 violent delinquency scale. For the wave 3
serious delinquency scale, only the genetic risk × parent–child relations interaction term is statistically significant. Genetic risk and
parent–child relations fail to have a statistically significant main effect
on serious delinquency. For the wave 3 violent delinquency scale, a
slightly different set of findings emerge. In the first model, the genetic

Table 3
The effects of genetic risk, parent–child relations, and gene × environment interactions on the wave 2 delinquency scales.
Wave 2 Serious delinquency scale
b
Genetic risk
Parent–child relations
Genetic risk × parent–child relations
Age
n
Notes:
Negative binomial regression equations;
Huber/White standard errors estimated.

SE

Wave 2 Violent delinquency scale

P

b

SE

P

b

P

b

SE

P

0.00
− 0.09
0.30
0.15

0.12
0.22
0.16
0.08
148

0.989
0.662
0.067
0.076

0.07
− 0.03

0.11
0.23

0.551
0.893

0.07

0.10
149

0.509

0.00
−0.09
0.35
0.08

0.12
0.21
0.16
0.09
149

0.999
0.688
0.023
0.371

0.05
− 0.09

0.12
0.23

0.667
0.711

0.12

0.09
148

0.151

SE

K.M. Beaver et al. / Psychiatry Research 175 (2010) 160–164

163

Table 4
The effects of genetic risk, parent–child relations, and gene × environment interactions on the wave 3 delinquency scales.
Wave 3 Serious delinquency scale
b
Genetic risk
Parent–child relations
Genetic risk × parent–child relations
Age
n

SE

Wave 3 Violent delinquency scale

P

b

SE

P

0.08
−0.23
0.56
−0.06

0.10
0.27
0.21
0.08
151

0.432
0.398
0.007
0.441

0.10
−0.12

0.10
0.24

0.360
0.612

− 0.12

0.08
151

0.137

b

P

b

SE

P

0.31
0.06

SE
0.14
0.29

0.026
0.830

− 0.19

0.10
153

0.054

0.23
−0.25
0.59
− 0.12

0.11
0.36
0.19
0.09
153

0.046
0.492
0.002
0.218

Notes:
Negative binomial regression equations;
Huber/White standard errors estimated.

they do not provide any information about which genes are influential.
Molecular genetic studies, in contrast, provide valuable information
about the specific genes that are associated with different psychopathologies. The genetic risk scale used in the current analysis culls
information from both of these lines of inquiry and provides a unique
measure of genetic risk that directly models polygenic effects.
Despite the potential strengths of a genetic risk scale, there are a
number of issues with it that need to be addressed in future studies. First,
exactly which genetic polymorphisms should be included in the genetic
risk index is not entirely clear. Ideally researchers would construct
genetic risk indexes by choosing genetic polymorphisms that have
consistently been linked to the phenotype being studied. However,
genetic variants, including some of the ones used in the current study,
are often found to be associated with a phenotype in one study, but not
in replication studies. Deciding whether these genes should be included
in a genetic risk index is not straightforward and future research will
need to explore the various ways that including or not including genes
impacts the effects of the genetic risk index. Second, it is possible that
the genes included in the current genetic risk index not only interact
with the environment, but also interact with each other (Beaver et al.,
2007). We were unable to test this possibility in the current study, but
future research would benefit by exploring the ways in which these
types of interaction effects could be modeled when using a genetic risk
index. Last, examining genetic effects with a genetic risk index is not
meant to replace the more conventional means of studying the effects of
certain genetic polymorphisms independently. A genetic risk index,
however, does represent one way of exploring the gene × environmental
foundations to polygenic phenotypes.
Although this study adds to a body of research revealing that
antisocial behaviors are partially the result of genetic factors, the
findings should be tempered in light of a number of limitations. Of
particular importance is that all of the antisocial phenotype measures
were based on self-reports, not official reports. This necessarily raises
the possibility that the measures could be unreliable. It should be noted,
however, that self-reports of criminal and delinquent involvement are
used frequently and have been shown to be valid and reliable
(Farrington et al., 1996; Jolliffe et al., 2003). Another limitation is that
the sample size used in the current study is relatively low (N = 145 to

risk measure has a statistically significant and positive effect on
violent delinquency. What this means is that African-American males
with more risk alleles report engaging in more acts of violence. As
with all the other models, the genetic risk × parent–child relations
interaction term is a significant predictor of violent delinquency
assessed at wave 3.
As Table 5 shows, the last set of models examines the predictors of
the number of police contacts variable and predictors of the composite
antisocial behavior index. For the number of police contact variable,
genetic risk and parent–child relations fail to have an effect. However,
consistent with all of the previous models, the genetic risk× parent–
child relations measure has a significant positive effect on the dependent
variable. The exact same findings are evident for the models predicting
the composite antisocial behavior index, with the genetic risk× parent–
child relations interaction emerging as the only significant predictor.
4. Discussion
The current study examined whether eight different antisocial
phenotypes were explained by gene × environment interactions. Unlike
most extant gene× environment research that examines the effect that a
particular gene has on phenotypic variation, we developed a unique
measure of genetic risk that was based on five different genetic
polymorphisms. This modeling strategy directly estimates polygenic
effects on antisocial behaviors. To explore whether the effect of genetic
risk was moderated by environmental stimuli, a composite measure of
parent–child relations was also employed. The results of the multivariate models revealed that genetic risk did not have any main effects
on antisocial behavior. Consistent with prior research (Jaffee et al.,
2005), however, the genetic risk scale did interact with the parent–child
relations scale to predict a significant amount of variance in all eight of
the antisocial phenotypes.
The composite measure of genetic risk represents a mid-point
between twin studies that model genetic effects as latent factors and
molecular genetic studies that typically examine the effects that one
particular gene has on an outcome measure. Genetic effects garnered
from twin studies are important as they provide specific information
about the percentage of variance accounted for by genetic factors, but

Table 5
The effects of genetic risk, parent–child relations, and gene × environment interactions on number of police contacts and the composite antisocial behavior index.
Number of police contacts
b
Genetic risk
Parent–child relations
Genetic risk × parent–child relations
Age
n
Notes:
Negative binomial regression equations;
Huber/White standard errors estimated.

SE

Composite Antisocial Behavior Index

P

b

SE

P

b

P

b

SE

P

0.05
0.19
0.21
0.02

0.12
0.12
0.10
0.09
159

0.687
0.112
0.035
0.860

0.17
0.30

0.15
0.31

0.247
0.332

0.06

0.13
145

0.641

0.10
0.04
0.50
0.07

0.19
0.35
0.19
0.14
145

0.594
0.919
0.009
0.591

0.08
0.23

0.12
0.15

0.477
0.113

0.00

0.09
159

0.974

SE

164

K.M. Beaver et al. / Psychiatry Research 175 (2010) 160–164

159). Even so, a sample of this size is large enough to support the
multivariate models calculated. Last, all of the genetic polymorphisms
included in the genetic risk index were given equal weighting. Whether
the alleles for some polymorphisms should be differentially weighted to
reflect stronger associations with antisocial phenotype remains an open
empirical question awaiting future research.
The analyses for this study were confined only to African-American
males. To our knowledge, this is the first time that a gene × environment
interaction effect on antisocial behaviors has ever been reported for
African-Americans. Given that African-Americans are disproportionately
overrepresented in the criminal justice system (DeLisi and Regoli, 1999),
much more gene× environment research with ethnic minorities is
needed. We should note, too, that these same models were calculated for
Caucasian males, but the results of the models did not reveal any
statistically significant genetic or gene ×environment effects. Although
speculative, these divergent results could be due to the fact that the allelic
frequencies of the genes included in this study are known to vary
considerably across racial and ethnic groups (Chang et al., 1996;
Gelernter et al., 1997, 1998). It should also be noted that parent–child
relations may differentially affect genetic expression in certain racial
groups. Future research needs to explore these possibilities in greater
detail as a way of unpacking the potential gene× environmental effects
that may be able to explain the wide racial gap in offending behaviors.
References
Arseneault, L., Moffitt, T.E., Caspi, A., Taylor, A., Rijsdijk, F.V., Jaffee, S.R., Ablow, J.C.,
Measelle, J.R., 2003. Strong genetic effects on cross-situational antisocial behaviour
among 5-year-old children according to mothers, teachers, examiner-observers,
and twins' self-reports. Journal of Child Psychology and Psychiatry 44, 832–848.
Bakermans-Kranenburg, M.J., van IJzendoorn, M.H., 2006. Gene–environment interaction
of the dopamine D4 receptor (DRD4) and observed maternal insensitivity predicting
externalizing behavior in preschoolers. Developmental Psychobiology 48, 406–409.
Beaver, K.M., Wright, J.P., DeLisi, M., Walsh, A., Vaughn, M.G., Boisvert, D., Vaske, J., 2007.
A gene × gene interaction between DRD2 and DRD4 is associated with conduct
disorder and antisocial behavior in males. Behavioral and Brain Functions 3, 30.
Cadoret, R.J., Langbehn, D., Caspers, K., Troughton, E.P., Yucuis, R., Sandhu, H.K., Philibert, R.,
2003. Associations of the serotonin transporter promoter polymorphism with
aggressivity, attention deficit, and conduct disorder in an adoptee population.
Comprehensive Psychiatry 44, 88–101.
Caspi, A., McClay, J., Moffitt, T.E., Mill, J., Martin, J., Craig, I.W., Taylor, A., Poulton, R., 2002.
Role of genotype in the cycle of violence in maltreated children. Science 297, 851–854.
Chang, F.M., Kidd, J.R., Livak, K.J., Pakstis, A.J., Kidd, K.K., 1996. The world-wide distribution
of allele frequencies at the human D4 receptor locus. Human Genetics 98, 91–101.
DeLisi, M., Regoli, B., 1999. Race, conventional crime, and criminal justice: the declining
importance of skin color. Journal of Criminal Justice 27, 549–557.
Farrington, D., Loeber, R., Stouthamer-Loeber, M., VanKammen, W.B., Schmidt, L., 1996.
Self-reported delinquency and a combined delinquency seriousness scale based on
boys, mothers, and teachers: concurrent and predictive validity for AfricanAmericans and Caucasians. Criminology 34, 493–517.
Frazzetto, G., Di Lorenzo, G., Carola, V., Proietti, L., Sokolowska, E., Siracusano, A., Gross, C.,
Troisi, A., 2007. Early trauma and increased risk for physical aggression during
adulthood: the moderating role of MAOA genotype. Plos One 2, e486.
Foley, D.L., Eaves, L.J., Wormley, B., Silberg, J.L., Maes, H.H., Kuhn, J., Riley, B., 2004.
Childhood adversity, monoamine oxidase A genotype, and risk for conduct disorder.
Archives of General Psychiatry 61, 738–744.
Gelernter, J., Kranzler, H., Cubells, J.F., 1997. Serotonin transporter protein (SLC6A4)
allele and haplotype frequencies and linkage disequilibria in African- and
European-American and Japanese populations and in alcohol-dependent subjects.
Human Genetics 101, 243–246.
Gelernter, J., Kranzler, H., Cubells, J.F., Ichinose, H., Nagatsu, T., 1998. DRD2 allele
frequencies and linkage disequilibria, including the -141CIns/Del promoter
polymorphism, in European-American, African-American, and Japanese subjects.
Genomics 51, 21–26.
Guo, G., Roettger, M.E., Shih, J.C., 2007a. Contributions of the DAT1 and DRD2 genes to
serious and violent delinquency among adolescents and young adults. Human
Genetics 121, 125–136.

Guo, G., Wilhelmsen, K., Hamilton, N., 2007b. Gene-lifecourse interaction for alcohol
consumption in adolescence and young adulthood: five monoamine genes.
American Journal of Medical Genetics Part B (Neuropsychiatric Genetics) 144B,
417–423.
Haberstick, B.C., Lessem, J.M., Hopfer, C.J., Smolen, A., Ehringer, M.A., Timberlake, D., Hewitt,
J.K., 2005. Monoamine oxidase A (MAOA) and antisocial behaviors in the presence of
childhood and adolescent maltreatment. American Journal of Medical Genetics 135B,
59–64.
Harris, K.M., Florey, F., Tabor, J., Bearman, P.S., Jones, J., Udry, J.R., 2003. The National
Longitudinal Study of Adolescent Health: Research Design. [www document].
URL: http://www.cpc.unc.edu/projects/addhealth/design.
Haynie, D.L., 2001. Delinquent peers revisited: does network structure matter?
American Journal of Sociology 106, 1013–1057.
Hopfer, C.J., Timberlake, D., Haberstick, B., Lessem, J.M., Ehringer, M.A., Smolen, A.,
Hewitt, J.K., 2005. Genetic influences on quantity of alcohol consumed by
adolescents and young adults. Drug and Alcohol Dependence 78, 187–193.
Huizinga, D., Haberstick, B.C., Smolen, A., Menard, S., Young, S.E., Corley, R.P., Stallings, M.C.,
Grotpeter, J., Hewitt, J.K., 2006. Childhood maltreatment, subsequent antisocial
behavior, and the role of monoamine oxidase A genotype. Biological Psychiatry 60,
677–683.
Jaffee, S.R., Caspi, A., Moffitt, T.E., Dodge, K.A., Rutter, M., Taylor, A., Tully, L.A., 2005.
Nature × nurture: genetic vulnerabilities interact with physical maltreatment to
promote conduct problems. Developmental Psychopathology 17, 67–84.
Jolliffe, D., Farrington, D.P., Hawkins, D.P., Catalano, R.F., Hill, K.G., Kosterman, R., 2003.
Predictive, concurrent, prospective and retrospective validity of self-reported
delinquency. Criminal Behaviour and Mental Health 13, 179–197.
Kendler, K.S., Kessler, R.C., Walters, E.E., MacClean, C., Neale, M.C., Heath, A.C., Eaves, L.J.,
1995. Stressful life events, genetic liability, and onset of an episode of major
depression in women. American Journal of Psychiatry 152, 833–842.
Kim-Cohen, J., Caspi, A., Taylor, A., Williams, B., Newcombe, R., Craig, I.W., Moffitt, T.E., 2006.
MAOA, maltreatment, and gene–environment interaction predicting children's mental
health: new evidence and a meta-analysis. Molecular Psychiatry 11, 903–913.
Mason, D.A., Frick, P.J., 1994. The heritability of antisocial behavior: a meta-analysis of twin
and adoption studies. Journal of Psychopathology and Behavioral Assessment 16,
301–323.
Miles, D.R., Carey, G., 1997. Genetic and environmental architecture of human
aggression. Journal of Personality and Social Psychology 72, 207–217.
Moffitt, T.E., 2005. The new look of behavioral genetics in developmental psychopathology:
gene–environment interplay in antisocial behaviors. Psychological Bulletin 131,
533–554.
Nilsson, K.W., Sjoberg, R.L., Damberg, M., Alm, P.O., Ohrvik, J., Leppert, J., Linkdstrom, L.,
Oreland, L., 2005. Role of the serotonin transporter gene and family function in
adolescent alcohol consumption. Alcoholism: Clinical and Experimental Research 29,
564–570.
Resnick, M.D., Bearman, P.S., Blum, R.W., Bauman, K.E., Harris, K.M., Jones, J., Tabor, J.,
Beuhring, T., Sieving, R.E., Shew, M., Ireland, M., Bearinger, L.H., Udry, J.R., 1997.
Protecting adolescents from harm: findings from the National Longitudinal Study of
Adolescent Health. Journal of the American Medical Association 278, 823–832.
Rhee, S.H., Waldman, I.D., 2002. Genetic and environmental influences on antisocial
behavior: a meta-analysis of twin and adoption studies. Psychological Bulletin 128,
490–529.
Ritchey, F.J., 2000. The Statistical Imagination: Elementary Statistics for the Social
Sciences. McGraw-Hill, New York.
Rutter, M., 2006. Genes and Behavior: Nature–nurture Interplay Explained. Blackwell,
Malden, MA.
Schreck, C.J., Fisher, B.S., Miller, J.M., 2004. The social context of violent victimization: a
study of the delinquent peer effect. Justice Quarterly 21, 23–47.
Tuvblad, C., Grann, M., Lichtenstein, P., 2006. Heritability for adolescent antisocial
behavior differs with socioeconomic status: gene–environment interaction. Journal
of Child Psychological and Psychiatry 47, 734–743.
Udry, J.R., 2003. The National Longitudinal Study of Adolescent Health (Add Health),
Waves I and II, 1994–1996; Wave III, 2001–2002 [machine-readable data file and
documentation]. Carolina Population Center, University of North Carolina at Chapel
Hill, Chapel Hill, NC.
Widom, C.S., Brzustowicz, L.M., 2006. MAOA and the “cycle of violence:” Childhood
abuse and neglect, MAOA genotype, and risk for violent and antisocial behavior.
Biological Psychiatry 60, 684–689.
Young, S.E., Smolen, A., Hewitt, J.K., Haberstick, B.C., Stallings, M.C., Corley, R.P., Crowley, T.J.,
2006. Interaction between MAO-A genotype and maltreatment in the risk for conduct
disorder: failure to confirm in adolescent patients. American Journal of Psychiatry 163,
1019–1025.


Related documents


PDF Document genetic risk african american
PDF Document genetic influence adoptees
PDF Document 1 s2 0 s0191886902003628 main
PDF Document perspectives on psychological science 2015 ferguson 646 66
PDF Document monoamine oxidase a regulates
PDF Document ese 691 week 3 assignment single subject


Related keywords