Fear conditioning during specific conditions in antisocial adolescents: a neuroimaging study.

In psychiatry, youngsters displaying a pattern of persistent and severe
antisocial behaviour are diagnosed with a disruptive behaviour disorder (DBD),
which is the most frequently seen juvenile psychiatric disorder in general
mental health clinics. In general populations the prevalence rates of DBD vary
around 5%, of which about 30% develop a Antisocial Personality Disorder in
adulthood. As such, it would be especially relevant to distinguish factors
explaining the onset of antisocial behavior and help inform interventions to
change this behavior.
One of the mechanisms hypothesized to drive antisocial behaviour development is
impaired fear conditioning. Impaired fear conditioning is thought to hamper
moral conscience development and therefore to increase the probability of
antisocial behaviour development. In line with this theory, studies have shown
impaired fear conditioning in adult antisocial individuals, relative to
controls.
A second neurobiological mechanism that has been proposed to increase the
probability of antisocial behaviour development, is dysfunctional reward /
punishment sensitivity. As antisocial individuals tend to choose immediate
rewarding antisocial strategies and disregard long-term negative outcomes, it
has been argued that they may be more sensitive to reward and less sensitive to
punishment than healthy controls. Dysfunctional reward/punishment sensitivity
may lead toincreased risk taking behaviour. Moreover, not only the sensitivity
to reward and punishment influences a child*s behaviour, also the flexibility
to change behaviour when rewarding or punishing contingencies are changing
(i.e. reversal learning) is of importance in this respect. Problems with
reversal learning causes inappropriate perseveration in formerly rewarding
behaviour even if this behaviour in a new condition is non-rewarding. Besides
inappropriate behaviour, problems in reversal learning in children may lead to
frustration and may result in reactive aggression, one manifestation of
antisocial behaviour.
To date, only one study investigated fear conditioning in adolescents (age
between 14 and 18 years) displaying antisocial behaviour and indeed showed
impaired conditioning in this sample. Moreover, although new neuroimaging
techniques have elucidated the brain areas and functions involved in fear
conditioning in healthy populations and antisocial adult populations, only one
study, of our own group, is using fMRI to study fear conditioning in antisocial
adolescents at this moment. Furthermore, there have been no studies
investigating whether disruptions in reward/punishment sensitivity and
(reversal) fear conditioning are malleable to intervention. The relation
between reward/punishment sensitivity and the degree of risk taking behaviour
has not been studied yet (See protocol §1).
Therefore, the current study aims to investigate the effect of three
experimental conditions (a basic condition, a methylphenidate condition and a
placebo condition) on some of the known neurobiological mechanisms linked to
the development of antisocial behaviour problems (e.g. hampered fear
conditioning, higher sensitivity for reward, lower sensitivity for punishment
and reversal problems) in DBD adolescents. To examine the effect of these
interventions on the neural substrates (function of relevant brain areas,
connectivity between these areas and their structure) of fear conditioning and
reward/punishment sensitivity, a functional neuroimaging study is proposed.

In the original protocol 25 DBD patients would complete the protocol in the
basic condition. However, due to slower than expected recruitment numbers of
DBD adolescents and a change in the availability of the MRI scanner, in
September 2014 (exact date after a positive decision of the CCMO has been
given) randomisation will take place over two (placebo and methylphenidate)
instead of three conditions (no-intervention, placebo and methylphenidate).
Moreover, adding a reversal phase to a fear conditioning task gives the
opportunity to test two different neurobiological characteristics of antisocial
behaviour in one paradigm.

The following hypotheses will be tested:
1. DBD adolescents will show diminished fear conditioning compared to HCs in
the basic condition, and this phenomenon is associated with altered function,
connectivity and structure of brain areas known to be involved in fear
conditioning.
2. DBD adolescents will show diminished fear reversal compared to HCs in the
basic condition.
3. DBD adolescents will show more sensitivity to reward and less sensitivity to
punishment compared to HCs in the basic condition, and these differences are
related to functional differences in brain activation patterns.
4. In DBD adolescents, fear conditioning and fear reversal will be potentiated
in the methylphenidate condition compared to the placebo and basic condition.
5. In DBD adolescents, methylphenidate will increase sensitivity to punishment
and decrease sensitivity to reward compared to the placebo condition and the
basic condition.
6. DBD adolescents will show higher levels of risk taking behaviour compared to
HCs.
7. High levels of risk taking behaviour will be correlated with increased
sensitivity to reward and decreased sensitivity to punishment for DBD
adolescents and HCs in the basic condition.

The primary objective is to investigate brain functioning during (reversal)
fear conditioning and reward/punishment conditions during different
experimental conditions. Secondary objectives include investigation of
connectivity between brain structures known to be involved in fear conditioning
and reward/ punishment using Diffusion Tensor Imaging (DTI) and fMRI techniques
(e.g. amygdala and ventromedial prefrontal cortex), as well as the study of
structural brain differences between adolescent DBD patients and HCs using
structural MRI. Additionally, this study aims to assess the relation between
reward/punishment sensitivity and risk taking behaviour.

This study is a non-therapeutic experimental study in a group of subjects with
a clinical diagnosis DBD and in a group of healthy control subjects. DBD
subjects will be studied during specific experimental conditions. Before the
MRI protocol starts, data will be collected concerning psychiatricdiagnoses,
psychological functioning and social functioning. Interviews, structered and
semi-structured, as well as questionnaires will be used for collecting data
from the adolescent and the primary caregiver. Risk taking behaviour will be
measured through a computerized task.. A total of 75 DBD patients (mean age 15
years) will be randomized over 3 different conditions (see protocol flowchart
page 20). The conditions consist of a basic condition, a methylphenidate
condition and a placebo condition. The healthy controls will perform the
protocol in the basic condition (see protocol §5.2). In the original protocol
25 DBD patients would complete the protocol in the basic condition. However,
due to slower than expected recruitment numbers of DBD adolescents and a change
in the availability of the MRI scanner, in September 2014 (exact date after a
positive decision of the CCMO has been given) randomisation will take place
over two (placebo and methylphenidate) instead of three conditions
(no-intervention, placebo and methylphenidate). Moreover, adding a reversal
phase to a fear conditioning task gives the opportunity to test two different
neurobiological characteristics of antisocial behaviour in one paradigm.

one group (n=25) receives Methylphenidate (0.3-0.4mg/kg) in capsules of 5mg.
one group (n=25) receives a placebo in the same amound of capsules as the
Methylphenidate group

DBD participants will be visited once at their home to be interviewed and to
fill in questionnaires. This house visits will last for 2 hours . On a second
appointment, the scanning session takes place (2 hours and 15 minutes).
For HCs 1 home visit will suffice for the interview and for filling in
questionnaires. This may last less than 2 hours, because HCs will report less
behavioral problems so the interview will last less compared to the interview
of DBD adolescents. The scanning session will take 2 hours and 15 minutes.
As the only injuries reported thus far in MRI-scanners have been attributed to
inadvertent presence of ferromagnetic materials or cardiac pacemakers, the risk
of the magnetic fields in the MRI during proper use (which includes screening
for these materials prior to scanning) is negligible according to current
scientific opinion. The noise in MRI-scanners may be quite loud and could
theoretically cause mild temporary hearing loss. The use of disposable earplugs
has been shown to provide a sufficient decrease in acoustic noise capable of
preventing this potential temporary hearing loss, while an additional
head-phone is used on top of these earplugs in our protocol, further reducing
the noise. Injury from MRI-procedures in our study is therefore not expected
(see protocol page 9).
To minimize the burden of the scanning procedure, the maximum duration of
scanning is limited to fifty minutes. Private correspondence with other
researchers using MRI in children suggests that 15 year-olds generally will
tolerate procedures lasting an hour. In one study, 16 adolescents (mean age
12.8) with DSM-IV diagnoses of anxiety disorders, who one would expect to be
troubled easily by a fMRI procedure, all tolerated and completed a fMRI fear
conditioning procedure. Furthermore, as the code of conduct concerning
non-therapeutic research in minors with MRI (as stated by the CCMO) requires,
participants will be given the option to practice in a mock scanner first,
reducing stress by accommodating them to the environment.
The use of a US may cause mild discomfort to the participants, due to its
inherent aversive nature. The US is already used in the ongoing study of M.
Cohn (ABR 28844) using predominantly the same imaging procotol as described
here, with approval of the METc of the VUmc. We have investigated alternative
modalities for a US and have found that an electric stimulus was the only
reliable US for use in our fMRI procedure. For example, an aversive sound burst
is difficult to use as US during fMRI scanning because of background noise and
may cause hearing damage due to the impossibility to fine tune soundburst
volume. An electric stimulus is the most commonly used US in the fear
conditioning literature, and is likely to be the most valid and robust US in
fMRI fear conditioning paradigms compared to other modalities. This US is
usually tolerated well and does not cause any injuries. In our laboratory site
there are researchers and technical crew experienced in fMRI fear conditioning
paradigms and the use of associated equipment, ensuring its proper use. To
minimize the burden of its use, the intensity will be individually calibrated
to a *unpleasant, but painless* level (see protocol §5.2). The US will be
delivered to the ankle instead of more threatening or sensitive areas. Also,
the use of a partial reinforcement strategy will reduce the number of exposures
to the US to seven.
After proper screening for contra indications, the risk of serious adverse
events associated with the use of methylphenidate is negligible. Physical
examination (blood pressure, heart rate and auscultation of the heart) will be
conducted by the principal investigator, a medical doctor experienced in
prescribing Methylphenidate to adolescents. Recently, large studies in children
and adolescents have shown that the use of stimulants is not associated with an
increased risk of cardiovascular incidents, neither during nor during the
period afterwards. In addition, it is likely that many DBD participants are
already familiar with Methylphenidate. Methylphenidate is a frequently
prescribed medication in youngsters with ODD or CD, with our without ADHD, and
is proven effective in the reduction of behavioral problems. It*s therefore
very likely that participants in the proposed study already use
methylphenidate, have used methylphenidate or will be advised to use
methylphenidate in the near future. In general, methylphenidate use has no or
only mild and transient side effects, side effects disappear after
discontinuation of the treatment (approximately four hours). Studies have shown
that a single dose of 0.5-0.6 mg/kg is well tolerated (see protocol page 9).