106 Biological Bases of Personality
Short time periods of prediction may confound en-
vironmental-developmental interactions that could mask the
influence of endogenous levels of T. Windle and Windle
(1995), in a retrospective longitudinal study, examined the
adult levels of plasma T in four groups: (a) those who were
aggressive only in childhood; (b) those who became aggres-
sive as adults; (c) those who were aggressive in both child-
hood and adulthood (continuity); and (d) those who were low
in aggression in childhood and adulthood. Adult onset and
continuity (in aggressiveness) groups had higher T levels as
adults than the other two groups. Other than aggressiveness,
the high-T adult groups had higher rates of antisocial person-
ality and a history of various signs of antisocial behavior.
Was the high level of T in these groups a product of their his-
tory or a sign of an earlier level of T that affected the devel-
opment of these behaviors? The authors admit that it is
impossible to answer this question.
High levels of cortisol are associated with stress and inhi-
bition and low levels with impulsivity and sensation seeking,
as noted previously. In baboons dominant and aggressive
males usually have low levels of cortisol and subordinate and
nonaggressive primates have higher levels of cortisol. As
with testosterone, cortisol varies considerably with recent
and long-term patterns of experience such as winning or los-
ing in fights. Low levels of cortisol have been found in psy-
chopathic, violent offenders (Virkkunen, 1985), but high
levels of cortisol are positively associated with hostility as
measured by hostility questionnaires (Keltikangas, Räikkö-
nen, & Adlercreutz, 1997; Pope & Smith, 1991). Chronic
feelings of hostility are often associated with anxiety and de-
pression, but the type of impulsive aggression seen in antiso-
cial personality represents a brief state of anger in a generally
unemotional personality.
Genetics
Behavior genetic studies of general hostility scales or ag-
gression in children have shown significant heritabilities.
However, it is possible that some aspects of hostility or ag-
gression may be more heritable than others. A twin study of
adult males using the BDHS revealed heritabilities ranging
from 28% for verbal hostility to 47% for assault (Coccaro,
Bergeman, Kavoussi, & Seroczynski, 1997). Verbal hostility
is the most common form and yet it had the least heritability
and the strongest environmental influence. An analysis of the
genetic influence on the correlations among the scales that
the assault scale had different underlying influences than the
other scales which shared a common genetic influence. With
the exception of the assault scale the genetic influence un-
derlying the scales is of a nonadditive type suggesting
Mendelian dominant or recessive or epistatic mechanisms. If
it is the former, there is the likelihood of finding a gene of
major effect in the general trait of aggression, apart from
physical assault type.
The MAO type-A gene has become a likely candidate for
this trait. Aggression in male mice is heightened by deletion
of the MAO-A gene (Cases et al., 1995), and a mutation in
the gene in a large Dutch family has been linked to mild re-
tardation and impulsive aggressive behavior (Bruner, Nelsen,
Breakfield, Ropers, & van Oost, 1993). The mutation is rare,
but the gene has a wide range of alleles varying in repeat
length. Subjects with one form, in contrast to those with
another form, had lower scores on an index of aggression/
impulsivity and the Barratt impulsiveness scale (Manuck,
Flory, Ferrell, Mann, & Muldoon, 2000). The life history of
aggression only approached significance and the BDHS did
not show significant differences between allele groups. Ap-
parently, the impulsivity was more salient than the aggres-
siveness in the combination. Consistent with the association
between low serotonin and aggression in the finding that the
allele group with the higher impulsive aggression score also
showed less response to a serotonergic challenge test.
Just as the findings on the MAO-A gene suggest one
source of the link between serotonin and aggression, another
gene has been found that suggests a genetic mechanism for
the association of norepinephrine with aggression. The
adrenergic-2A receptor gene (ADRA2A) plays a role in mod-
ulating norepinephrine release in the locus coeruleus. Alleles
of this gene were associated with scales for hostility and im-
pulsivity in a younger student sample and impulsivity alone
in an older sample (Comings et al., 2000).SummaryExtreme violence has been associated with EEG evidence of
cortical abnormality usually in the form of an excess of slow
wave activity (underarousal) or focalized EEG abnormalities
in the temporal lobes. Brain scans have confirmed the tem-
poral lobe abnormalities and also found an equal incidence
of frontal lobe abnormalities. A reduced P300 cortical EP re-
sponse has also been found in prisoners with a history of ex-
tremely violent behavior. The reduced activity and reactivity
in the frontal lobes may reflect a deficit in inhibitory capac-
ity, which is part of the executive function of these lobes.
The abnormal activity of the temporal lobe may be sympto-
matic of abnormal amygdala function because this lobe is in
close proximity to the underlying amygdala. An MRI study
has revealed temporal lobe lesions in about one third of vio-
lent patients. Hostility or anger proneness is related to a high
level of cardiovascular, noradrenergic, and testosterone and