Cognitive Control and Comorbid Methamphetamine Abuse in Schizophrenia Patients

1
Cognitive Control and Comorbid Methamphetamine
Abuse in Schizophrenia Patients R Salo1, S
Ursu1, TE Nordahl 1 , Y Natsuaki 1, M Leamon 1,
MH Buonocore 3 C.S. Carter1,2 Depts. of
1Psychiatry 2Psychology, and Radiology 3
University of California, Davis,
Imaging Results
Behavioral Results (2)
Introduction
Methods
ROIs were based on previous spectroscopy
findings of abnormalities in MA and SZ subjects
Incongruent Error Rates
Post Error Slowing (adjusted for RT differences)
Mean Difference of Beta Coefficients for
Incong-Cong trials (Averaged for each ACC ROI)
Rostral ACC (BA 32,10)
Caudal ACC (BA 32,24)
Baseline RT
Stroop RT Effect Incong-Cong-adj for RT diff
Group Age Edu Yrs Use
SZ 30.3 (5.9) 13.3 N/A
SZ MA 33.4 (8.6) 10.8 10.4 yrs
MA 34.8 (2.5) 12.0 12.3 yrs
CTL 32.0 (10.6) 15.1 N/A
Results Despite lack of group differences in
Incong Error rates, group differences emerged in
RT adjustment following error trials. No sig
group differences were observed in the Stroop
Effect.
Mean Difference of Beta Coefficients for
Incong-Cong trials (Averaged for each DLPFC ROI)
Left DLPFC (BA 9,46)
Right DLPFC (BA 9,46)
Imaging Parameters
Behavioral Results (1)
Type EPI Slices 36
Scanner 3T Siemens Trio TR 2000 ms
Voxel size 3.438 x 3.438 x 3.4mm TE 25 ms
Flip angle 90 FOV 220 mm
Trial to Trial Reaction Time Adjustments Rt to
Incong Trials following Incong Trials PreInc vs
Incong Trials following Congruent Trials PreCong
R
L
R
L
Controls
MA Abusers
Analysis of fMRI Imaging Data Statistical
single-subject correlation maps were calculated
for activity in specific regions of interest
ROIs including ACC and DLPFC. Covariates
were used in the analysis (congruent, incongruent
and error trials, the latter as a covariate of
non-interest). To generate the regressors used in
the statistical model, each variable of interest
was convolved with a double gamma hemodynamic
response function (Friston et al, 1995) These
regressors were entered simultaneously into a
general linear model implemented using AFNI
software (Cox, 1996). This procedure generated a
semipartial correlation maps for each subject,
subsequently used to compute differences between
correlation coefficients of conditions of
interest (e.g. incongruent vs. congruent, etc)
Controls vs SZ SZ MA
n.s.
p.07
STROOP STIMULI
Controls exhibited reduced RTs to I-I sequences
(p.07). In contrast, MA abusers, SZ or SZMA
failed to exhibit this pattern of cognitive
regulation.
Funded by NIDA