96 Biological Bases of Personality
Avoidance (HA) instead of the N or anxiety trait scales used
by other investigators. HA, however, is not a pure scale of the
N factor but lies between the E and N dimensions, constitut-
ing a measure of introverted neuroticism. It is defined in the
same way that Gray defines trait anxiety: a sensitivity to cues
associated with punishment and nonreward (frustration) and
a tendency to avoid them.
Gray’s (1982) model suggests that norepinephrine in the
dorsal ascending noradrenergic system (DANA) originating
in the locus coeruleus is the major neurotransmitter involved
in anxiety, although high levels of serotonin may mediate the
behavioral inhibition that is associated with high levels of
anxiety. Redmond (1977), from a psychiatric viewpoint, sees
the DANA as an alarm system at lower levels and a panic
provoker at high levels of activity. In contrast to these two
theorists, Cloninger, Svrakic, and Przybeck (1993) proposed
that high levels of serotonin activity underlie the trait of HA
whereas norepinephrine activity is related to another trait
called Reward Dependence.
In patients there has been little evidence of higher levels
of the norepinephrine metabolite MHPG in anxiety patients
compared to normals, although a more recent study by
Spivak et al. (1999) showed higher levels of MHPG in
plasma of patients with combat-related posttraumatic stress
disorder than in controls.
The alpha-2 receptor functions as a homeostatic regulator
of the norepinephrine systems, tuning them down when ex-
cessive neurotransmitter levels are detected in the synapse.
Yohimbine is a antagonist to this receptor and therefore po-
tentiates the activity of the norepinephrine system, just as a
broken thermostat results in an overheated room. Yohimbine
increases MHPG levels and provokes panic attacks in pa-
tients with panic disorders, although it does not have these ef-
fects in normal controls (Charney & Heninger, 1986).
Cameron et al. (1996) replicated a previous result finding a
decreased number of alpha-2 receptors in panic disorder. One
might extrapolate that MHPG should correlate with N or anx-
iety over the range in normals and other patient groups. How-
ever, as noted earlier, high N in normals may resemble GAD
more than panic disorder. Heinz, Weingarten, Hommer,
Wolkowitz, and Linnoila (1999) reported a high correlation
between CSF MHPG and an anxiety scale in a combined
group of abstinent alcoholics and normals. A stress resistant
group, defined by N and similar measures, had lower plasma
MHPG after a mild stressor than did a nonresistant (high-N)
group (de Leeuwe, Hentschel, Tavenier, & Edelbroek, 1992).
Norepinephrine may be one of the factors underlying N, but
it may be the dysregulation of norepinephrine by a lack of the
receptors needed for this and a consequent tendency to be un-
able to cope with stress, rather than the basal level of activity
in the norepinephrine system, which is related to N.
Cloninger’s biosocial theory of personality proposes that
the trait of harm avoidance is related to behavioral inhibition
mediated by serotonergic activity in the brain. Earlier studies
showed no correlation between between CSF levels of the
serotonin metabolite, 5-HIAA, and N. A more recent study
has found a positive correlation between CSF 5-HIAA and N
but in a sample of depressed patients (Roy, 1999). Constan-
tino and Murphy’s (1996) study of the prediction of infant
temperament from CSF levels of 5-HIAA showed no rela-
tionships between this metabolite and emotionality, sootha-
bility, or activity in infants.
Studies of normals using serotonin challenges, drugs that
stimulate serotonergic activity, and indirect measures of sero-
tonin response in normals have yielded mixed results includ-
ing both positive (Gerra et al., 2000; Hansenne & Ansseau,
1999), nonsignificant (Ruegg et al., 1997), and a negative re-
lationship (Mannuck et al., 1998) with N. The first three of
these studies used the HA scale, whereas the last used the
N scale, but with a much larger number of normal subjects
than in the other studies. Serotonin seems to be implicated in
harm avoidance, but the nature of that relationship is open
to question. As with other neurotransmitters, the personality-
relevant aspects of serotonin may have more to do with recep-
tor number and sensitivity than with basal levels of transmitter
activity.HormonesDaitzman and Zuckerman (1980) found that testosterone (T)
in males correlated negatively with various MMPI indexes of
anxiety, depression, and neuroticism; that is, subjects with
neurotic tendencies were low on T. Dabbs, Hopper, and
Jurkovic (1990) reported a significant negative correlation
between T and N in one study, but this was not replicated in
another larger study of males; and in an even larger study of
over 5,000 veterans T was not correlated with any MMPI in-
dexes of trait anxiety or N. In still another study Dabbs et al.
report significant negative correlations between T and a mea-
sure of pessimism in both males and females. T reflects both
trait and state characteristics; that is, it is affected by immedi-
ate stressful experiences, particularly those involving success
or defeat in competitive activities (Dabbs, 2000). The rela-
tionship with pessimism may reflect a history of defeat and
consequent expectations for future failures. This depressive
attitude may underlie negative relationships with N if any
such relationships do exist.
Cortisol is one of the end products of activation of the
hypothalamic-pituitary adrenocortical (HYPAC) system, a
stress-reactive hormonal system. Like T, cortisol reactivity
has both trait and state characteristics. Elevated cortisol is
associated with major depressive disorder as a trait but is