Neuroradiology (2014) 56:69–77 DOI 10.1007/s00234-013-1281-3
FUNCTIONAL NEURORADIOLOGY
Assessment of abstract reasoning abilities in alcohol-dependent subjects: an fMRI study Deepika Bagga & Namita Singh & Sadhana Singh & Shilpi Modi & Pawan Kumar & D. Bhattacharya & Mohan L. Garg & Subash Khushu
Received: 28 June 2013 / Accepted: 29 August 2013 / Published online: 13 November 2013 # Springer-Verlag Berlin Heidelberg 2013
Abstract Introduction Chronic alcohol abuse has been traditionally associated with impaired cognitive abilities. The deficits are most evident in higher order cognitive functions, such as abstract reasoning, problem solving and visuospatial processing. The present study sought to increase current understanding of the neuropsychological basis of poor abstract reasoning abilities in alcohol-dependent subjects using functional magnetic resonance imaging (fMRI). Methods An abstract reasoning task-based fMRI study was carried out on alcohol-dependent subjects (n =18) and healthy controls (n =18) to examine neural activation pattern. The study was carried out using a 3-T whole-body magnetic resonance scanner. Preprocessing and post processing was performed using SPM 8 software. Results Behavioral data indicated that alcohol-dependent subjects took more time than controls for performing the task but there was no significant difference in their response accuracy. Analysis of the fMRI data indicated that for solving abstract reasoning-based problems, alcohol-dependent subjects showed enhanced right frontoparietal neural activation
D. Bagga : N. Singh : S. Singh : S. Modi : P. Kumar : S. Khushu NMR Research Centre, Institute of Nuclear Medicine and Allied Sciences (INMAS), Timarpur, Delhi, India D. Bhattacharya Department of Psychiatry, Base Hospital, Delhi Cantt, India M. L. Garg Department of Biophysics, Panjab University, Chandigarh, India S. Khushu (*) NMR Research Centre, INMAS, DRDO, Lucknow Road Timarpur, Delhi, India e-mail: [email protected]
involving inferior frontal gyrus, post central gyrus, superior parietal lobule, and occipito-temporal gyrus. Conclusions The extensive activation observed in alcohol dependents as compared to controls suggests that alcohol dependents recruit additional brain areas to meet the behavioral demands for equivalent task performance. The results are consistent with previous fMRI studies suggesting decreased neural efficiency of relevant brain networks or compensatory mechanisms for the execution of task for showing an equivalent performance. Keywords Alcoholism . fMRI . Abstract reasoning . Brain . Functional
Introduction Brain damage is a common and potentially severe consequence of long-term, heavy alcohol consumption. There is considerable evidence that prolonged, excessive alcohol consumption results in neuropsychological deficits [1, 2]. In particular, abstract reasoning, visuospatial, and problem solving abilities seem to be frequently impaired after years of heavy alcohol consumption [3]. These are the cognitive operations linked to the frontal and the parietal cortex that guide complex behavior over time through planning, decisionmaking, and response control. Structural studies have shown diffuse bilateral cortical atrophy in alcohol-dependent subjects with the frontoparietal areas and cerebellum showing the earliest and most extensive shrinkage [4–6]. Thus, these studies add to the evidence that impairment in these cognitive functions is a characteristic sequela of chronic heavy drinking. Several studies of alcoholism have reported significant correlations between intellectual impairment and cerebral atrophy [7, 8]. Also, another study by Rogers et al. have shown that
70
impairments in motor functioning in alcoholics is attributed to diminished connectivity in fronto-cerebellar circuits [6]. Previous neuropsychological studies indicate that alcoholdependent subjects do show deficits in crystallized intelligence but these deficits are more subtle than are the deficits exhibited on fluid intelligence. The crystallized intelligence is dependent on acquired knowledge as it relies on accessing information from long-term memory whereas fluid intelligence is independent of acquired knowledge and is based on the capacity to think logically in novel situations. Abstract reasoning has been considered to be an important domain for assessment of fluid intelligence [9]. Impairments in fluid intelligence skills of visuospatial processing and abstract reasoning appear to reflect impairments in higher cognitive functions of perceptual analysis and synthesis [10]. Various functional connectivity studies have also highlighted the importance of frontoparietal control system in executive functioning [11]. The latter is also known as a task-positive system and is activated by a variety of demanding cognitive tasks like reasoning, attention, and decision-making [12]. Studies have also shown the role of frontal cortex in abstract reasoning. The connectivity of frontoparietal regions, as suggested by a recent study [13], further strengthens the involvement of frontoparietal cortex in abstract reasoning. Additionally, studies have also shown the importance of cortico-striato-thalamic circuits in cognitive processes of goal directed behavior and executive functioning. This circuit consists of looped neural pathways that connect the thalamus and striatum regions to the cerebral cortex, and connect the cerebral cortex back to these areas. The latter has been shown to be impaired in alcoholism [14]. The frontoparietal system in alcoholics might compensate for the impaired cortico-striato-thalamic circuits to support abstract reasoning abilities. Despite the fact that alcohol affects the abstract reasoning abilities to such a great extent, to date, little research has been done to assess the functional correlates associated with deficits in abstract reasoning abilities observed in alcohol dependents. The goal of our study was to investigate the functional correlates of impaired abstract reasoning abilities in alcoholdependent subjects. As abstract reasoning is a complex task that requires numerous cognitive operations such as working memory, attention, visuospatial processing, inhibitory control, decision-making, and abstraction, a matching baseline was designed to require similar processing in terms of visual encoding, decision process, and motor response execution, without the need for reasoning. Based on previous neuropsychological and functional magnetic resonance imaging (fMRI) studies, we hypothesized that alcohol-dependent subjects will show an enhanced activation bilaterally in frontal and frontoparietal brain areas. The rationale behind choosing a nonverbal abstract reasoning task was that these tasks are considered less biased than languagebased problems because of the use of symbols and pictures instead of concrete examples. To the best of our knowledge,
Neuroradiology (2014) 56:69–77
this is the first abstract reasoning-based fMRI study in alcohol-dependent subjects.
Methods Subjects The study included 18 alcohol-dependent subjects and 18 healthy individuals (DSM IV criteria). All study participants were men, between 30 and 50 years of age, and non-smokers. Alcohol-dependent subjects were recruited from a local hospital. The control subjects were recruited from the local community (see Table 1). As per DSM IV basic criteria, for alcohol abuse or alcohol dependence, impairments in physical, social, legal or any other area of role performance by an individual is required. Hence, recruitment of healthy individuals from community excluded individuals with alcohol dependence. Five control subjects out of 18 controls were occasional drinkers (20 g of pure ethanol/month). All subjects were also evaluated using the Alcohol Use Disorders Identification Test (AUDIT) [15]. AUDIT is a very reliable and simple screening tool which is sensitive to early detection of risky and high risk (or hazardous and harmful) drinking. All the subjects were right handed as assessed by Edinburgh Inventory [16]. Imaging studies were carried out at our institute. The inclusion criterion for alcohol-dependent patients was detoxification for at least 2 weeks and abstinence as assessed by normal levels of gamma glutamyl transferase (GGT) (Table 1). GGT test is widely used as a marker for alcohol intake. Elevated levels of GGT indicate excessive alcohol consumption. As our study subjects were recruited after detoxification of 2 weeks, they were not on any psychotropics at the time of fMRI. The exclusion criteria included (1) signs or symptoms of malnutrition, (2) signs of liver dysfunction: aspartate aminotransferase/ alanine aminotransferase ratio greater than 2 [17]. All the alcoholic patients were clinically normal with no past history of psychiatric disorders. The clinical assessment included detailed medical history, neurological, and neuropsychological examinations, and laboratory tests (routine hematology and biochemistry screen, thyroid function tests). All subjects underwent a thorough medical, psychiatric, and neurological examination by a senior psychiatrist. Exclusion criteria for all subjects included any kind of neurological symptoms (cerebellar, sensory, or motor dysfunctions), history of psychiatric disorder (other than alcohol dependence for patients), medical conditions that may alter cerebral function (i.e., cardiovascular, endocrinological, autoimmune, or oncological diseases), and brain trauma (seizures, degenerative disease, previous head injury with loss of consciousness). In addition to these examinations, the presence of visible abnormalities on T2 images (assessed by a radiologist), pacemaker, bypass surgery or metallic implants that would preclude MRI
Neuroradiology (2014) 56:69–77
71
Table 1 Characteristics of study groups Characteristics
Alcoholdependent subjects (n =18)
Healthy subjects (n =18)
Age(years) Body mass index (kg/m2) Education (years) AUDIT Alcohol consumptiona Duration of dependence (years) Abstinence (weeks) Age (years) at first drinking Age (years) at the onset of dependence Biological variables: g-Glutamyl-transferase Alanine aminotransferase (U/l) Aspartate aminotransferase (U/l) Aspartate aminotransferase/alanine aminotransferase
36.5±5 24.5±4.1 10±1.89 30.2±4.6 153.3±19.2 4.43±1.3
35.2±3.7 24.6±3.5 10±1.85 2±1.3b,c
Laboratory norms
17.47±4.39 23.7±3.1 33.2±6.2
51.3±2.1 26.4±14.1
≤53 ≤38
27.3±16.1
≤40
1.03±0.27
≤2
a
Consumption was defined as grams of pure alcohol per day during 3 months preceding detoxification
b p value (≤0.05) for between-group comparisons performed using twosample t test c
Average score of five control subjects who were occasional drinkers
scan or substance abuse (other than alcohol and tobacco) resulted to the exclusion of the subject. Therefore, the set of exclusion criteria led up to select alcohol-dependent patients without apparent neurological conditions. The local ethics committee approved the study and written informed consent was obtained from all participants after the procedures had been fully explained. Abstract reasoning task The abstract reasoning task was divided into two conditions: reasoning condition (R) and control condition (C). Each condition consisted of two phases: stimulus and response. Subjects were presented with the stimulus phase for 4 s, immediately followed by the response phase for 3.5 s. During the stimulus phase, subjects saw three pictures simultaneously in one row across the top of the screen. During the response phase, subjects saw two possible answer choices presented simultaneously in one row across the bottom of the screen. To select an answer, all subjects responded with a response grip in each of their hands. They pressed the button on the left to select the answer on the left, and the button on the right to select the answer on the right. During the trials, the phrase
“NEXT?” appeared at the top of the screen for the duration of the trial. During R trials, subjects saw three sequential stimuli presented in one row at the top of the screen. During C trials, subjects saw the same stimuli in all three positions of the top row. Subjects looked at the three stimuli and determined what the fourth pictures in the sequence would be (Fig. 1). During the response period, the top row disappeared, and subjects saw two possible answer choices on the bottom of the screen. The stimuli in the R condition consisted of line drawings of shapes such as circles, squares, and triangles. Sequences were designed such that they changed along only one dimension: number, position, shading, or size [18, 19]. Trials were designed such that there were an equal number of trial types that varied across each of the four dimensions. The answers for the R trials consisted of the correct response, and a foil that was either a picture in the presented sequence, or a variation along another dimension. C stimuli were created by randomly selecting one of the stimuli from the R sequences. As with the R trials, the answers for the C trials consisted of the correct response and a foil [19]. The control task was designed to require similar processing in terms of visual encoding, decision process, and motor response execution, without the need for reasoning. All subjects were first introduced to the tasks and response device in out-of-magnet training that included introduction and instruction slides as well as sample trials of Reasoning and Control trials on a laptop to make sure that they understood the task. Training was geared at thoroughly familiarizing participants with the modes of stimulus presentation, task requirements, and response options.
Scanning protocol The participants were scanned inside a 3-T whole-body MRI system (Magnetom Skyra, Siemens, Germany) equipped with a circularly polarized 20 channel matrix head and neck coil and 45 mT/m actively shielded gradient system. Subjects lay in the supine position with their heads supported and immobilized within the head coil using foam-pads (vendor provided), to minimize head movement and gradient noise. Thirty-six axial slices parallel to the bicommissural plane through the frontoparietal cortex covering the whole brain volume using gradient echo-based interleaved EPI sequence (matrix=64×64, field of view=210 mm, TE=36 ms, TR=3 s, flip angle=90°, slice thickness=3 mm, voxel size=3.28×3.28×3 mm3) were obtained. For anatomical reference, a T1-weighted 3D gradient echo sequence (MPRAGE: Magnetization Prepared Rapid Acquisition Gradient Echo, 160 sagittal slices, slice thickness= 1 mm, field of view=256 mm, TR=1900 ms, TE=2.07 ms) image data set was acquired coplanar with the functional scan, to allow for spatial registration of each subject’s data into a standard coordinate space.
72
Neuroradiology (2014) 56:69–77
Fig. 1 Experimental task. For all stimuli, subjects saw an instructional cue word and three pictures at the top of the screen. After 4 s, the top row disappeared, and two answer choices appeared at the bottom of the
screen. For both reasoning (R) and control (C) trials, the word “NEXT?” cued participants to determine the next picture in the sequence
fMRI Protocol
Data analysis
Block paradigm (BABABABABABAB) with alternating six phases of activation (A) and baseline (B) was chosen. 190 sequential image volumes (belonging to six alternating cycles+one baseline for eliminating T1 saturation effects and acclimatization of the patient to the gradient noise) were taken. The baseline (30 s) and the experimental task (60 s) blocks consisted of four and eight stimuli each. The total fMRI acquisition time was 9 min and 34 s including the activation (6 min) and baseline phases (3.5 min) along with an introductory screen (4 s) at the beginning of task. Stimuli were presented using fMRI hardware from NordicNeuroLab and the subject’s response was monitored with the help of Nordic response device system (NordicNeuroLab (http://www.nordicneurolab.com/Productsand Solutions/Nordic fMRI solution/index.aspx)). Timing of stimuli presentation was synchronized with the scanner image volume acquisition rate. Response times (RT) and response accuracies (RA) were recorded and stored on a PC outside the magnetic resonance scanner room, for off-line statistical analysis.
At first, fMRI data were transformed into NIFTI format. fMRI image data were processed with Statistical Parametric Mapping (SPM8, Wellcome Department of Cognitive Neurology, London, UK, http://www.fil.ion.ucl.ac.uk/spm) software package implemented in MATLAB R2008a (Version 7.6.0, Math works, Sherbon, MA). The first ten brain volumes (one baseline) of each fMRI data set were discarded to remove the initial transit signal fluctuations and subsequent images were realigned within the session to remove any minor movements. Translational or rotational movement of ±1.5 mm was not considered for analysis. The T1-weighted high-resolution anatomical images were co-registered with fMRI images and spatially normalized according to the Montreal Neurological Institute (MNI) brain template. The time-course images were normalized using the same normalization parameters and then smoothed with a 6×6×6 mm3 (full width at half maximum) Gaussian smoothing kernel. The EPI-images were high-pass filtered (128 s) to remove artifacts due to cardio respiratory and other cyclical influences. A statistic parametric map
Neuroradiology (2014) 56:69–77
(SPM) was generated for each subject under each condition by fitting the stimulation paradigm to the functional data, convolved with a hemodynamic response function. Conditionspecific effects at each voxel were estimated using the general linear model [20]. Individual first-level contrast images were generated for the abstract reasoning task versus baseline contrast (FWE-corrected, p 10 voxels for two-sample t test, which resulted in a corrected threshold of p alcohol dependents When subtracting the alcohol group from the control group, there remained no significant areas showing greater activation.
Results Drinking history variables for the alcohol-dependent subjects are presented in Table 1. The alcohol-dependent and control groups were similar with respect to mean age, years of education, and body mass index. Behavioral performance During the active phase (abstract reasoning trials), there was no difference with respect to accuracy (two-sample t test, p ≤0.12) between alcohol-dependent subjects (percentage accuracy: mean=63.52, SD=3.37; total no. of responses: mean=41.23,
Fig. 2 3D-rendered whole brain activation map for alcohol-dependent subjects. Activation maps are displayed at a threshold of t >10, p 10, p
FUNCTIONAL NEURORADIOLOGY
Assessment of abstract reasoning abilities in alcohol-dependent subjects: an fMRI study Deepika Bagga & Namita Singh & Sadhana Singh & Shilpi Modi & Pawan Kumar & D. Bhattacharya & Mohan L. Garg & Subash Khushu
Received: 28 June 2013 / Accepted: 29 August 2013 / Published online: 13 November 2013 # Springer-Verlag Berlin Heidelberg 2013
Abstract Introduction Chronic alcohol abuse has been traditionally associated with impaired cognitive abilities. The deficits are most evident in higher order cognitive functions, such as abstract reasoning, problem solving and visuospatial processing. The present study sought to increase current understanding of the neuropsychological basis of poor abstract reasoning abilities in alcohol-dependent subjects using functional magnetic resonance imaging (fMRI). Methods An abstract reasoning task-based fMRI study was carried out on alcohol-dependent subjects (n =18) and healthy controls (n =18) to examine neural activation pattern. The study was carried out using a 3-T whole-body magnetic resonance scanner. Preprocessing and post processing was performed using SPM 8 software. Results Behavioral data indicated that alcohol-dependent subjects took more time than controls for performing the task but there was no significant difference in their response accuracy. Analysis of the fMRI data indicated that for solving abstract reasoning-based problems, alcohol-dependent subjects showed enhanced right frontoparietal neural activation
D. Bagga : N. Singh : S. Singh : S. Modi : P. Kumar : S. Khushu NMR Research Centre, Institute of Nuclear Medicine and Allied Sciences (INMAS), Timarpur, Delhi, India D. Bhattacharya Department of Psychiatry, Base Hospital, Delhi Cantt, India M. L. Garg Department of Biophysics, Panjab University, Chandigarh, India S. Khushu (*) NMR Research Centre, INMAS, DRDO, Lucknow Road Timarpur, Delhi, India e-mail: [email protected]
involving inferior frontal gyrus, post central gyrus, superior parietal lobule, and occipito-temporal gyrus. Conclusions The extensive activation observed in alcohol dependents as compared to controls suggests that alcohol dependents recruit additional brain areas to meet the behavioral demands for equivalent task performance. The results are consistent with previous fMRI studies suggesting decreased neural efficiency of relevant brain networks or compensatory mechanisms for the execution of task for showing an equivalent performance. Keywords Alcoholism . fMRI . Abstract reasoning . Brain . Functional
Introduction Brain damage is a common and potentially severe consequence of long-term, heavy alcohol consumption. There is considerable evidence that prolonged, excessive alcohol consumption results in neuropsychological deficits [1, 2]. In particular, abstract reasoning, visuospatial, and problem solving abilities seem to be frequently impaired after years of heavy alcohol consumption [3]. These are the cognitive operations linked to the frontal and the parietal cortex that guide complex behavior over time through planning, decisionmaking, and response control. Structural studies have shown diffuse bilateral cortical atrophy in alcohol-dependent subjects with the frontoparietal areas and cerebellum showing the earliest and most extensive shrinkage [4–6]. Thus, these studies add to the evidence that impairment in these cognitive functions is a characteristic sequela of chronic heavy drinking. Several studies of alcoholism have reported significant correlations between intellectual impairment and cerebral atrophy [7, 8]. Also, another study by Rogers et al. have shown that
70
impairments in motor functioning in alcoholics is attributed to diminished connectivity in fronto-cerebellar circuits [6]. Previous neuropsychological studies indicate that alcoholdependent subjects do show deficits in crystallized intelligence but these deficits are more subtle than are the deficits exhibited on fluid intelligence. The crystallized intelligence is dependent on acquired knowledge as it relies on accessing information from long-term memory whereas fluid intelligence is independent of acquired knowledge and is based on the capacity to think logically in novel situations. Abstract reasoning has been considered to be an important domain for assessment of fluid intelligence [9]. Impairments in fluid intelligence skills of visuospatial processing and abstract reasoning appear to reflect impairments in higher cognitive functions of perceptual analysis and synthesis [10]. Various functional connectivity studies have also highlighted the importance of frontoparietal control system in executive functioning [11]. The latter is also known as a task-positive system and is activated by a variety of demanding cognitive tasks like reasoning, attention, and decision-making [12]. Studies have also shown the role of frontal cortex in abstract reasoning. The connectivity of frontoparietal regions, as suggested by a recent study [13], further strengthens the involvement of frontoparietal cortex in abstract reasoning. Additionally, studies have also shown the importance of cortico-striato-thalamic circuits in cognitive processes of goal directed behavior and executive functioning. This circuit consists of looped neural pathways that connect the thalamus and striatum regions to the cerebral cortex, and connect the cerebral cortex back to these areas. The latter has been shown to be impaired in alcoholism [14]. The frontoparietal system in alcoholics might compensate for the impaired cortico-striato-thalamic circuits to support abstract reasoning abilities. Despite the fact that alcohol affects the abstract reasoning abilities to such a great extent, to date, little research has been done to assess the functional correlates associated with deficits in abstract reasoning abilities observed in alcohol dependents. The goal of our study was to investigate the functional correlates of impaired abstract reasoning abilities in alcoholdependent subjects. As abstract reasoning is a complex task that requires numerous cognitive operations such as working memory, attention, visuospatial processing, inhibitory control, decision-making, and abstraction, a matching baseline was designed to require similar processing in terms of visual encoding, decision process, and motor response execution, without the need for reasoning. Based on previous neuropsychological and functional magnetic resonance imaging (fMRI) studies, we hypothesized that alcohol-dependent subjects will show an enhanced activation bilaterally in frontal and frontoparietal brain areas. The rationale behind choosing a nonverbal abstract reasoning task was that these tasks are considered less biased than languagebased problems because of the use of symbols and pictures instead of concrete examples. To the best of our knowledge,
Neuroradiology (2014) 56:69–77
this is the first abstract reasoning-based fMRI study in alcohol-dependent subjects.
Methods Subjects The study included 18 alcohol-dependent subjects and 18 healthy individuals (DSM IV criteria). All study participants were men, between 30 and 50 years of age, and non-smokers. Alcohol-dependent subjects were recruited from a local hospital. The control subjects were recruited from the local community (see Table 1). As per DSM IV basic criteria, for alcohol abuse or alcohol dependence, impairments in physical, social, legal or any other area of role performance by an individual is required. Hence, recruitment of healthy individuals from community excluded individuals with alcohol dependence. Five control subjects out of 18 controls were occasional drinkers (20 g of pure ethanol/month). All subjects were also evaluated using the Alcohol Use Disorders Identification Test (AUDIT) [15]. AUDIT is a very reliable and simple screening tool which is sensitive to early detection of risky and high risk (or hazardous and harmful) drinking. All the subjects were right handed as assessed by Edinburgh Inventory [16]. Imaging studies were carried out at our institute. The inclusion criterion for alcohol-dependent patients was detoxification for at least 2 weeks and abstinence as assessed by normal levels of gamma glutamyl transferase (GGT) (Table 1). GGT test is widely used as a marker for alcohol intake. Elevated levels of GGT indicate excessive alcohol consumption. As our study subjects were recruited after detoxification of 2 weeks, they were not on any psychotropics at the time of fMRI. The exclusion criteria included (1) signs or symptoms of malnutrition, (2) signs of liver dysfunction: aspartate aminotransferase/ alanine aminotransferase ratio greater than 2 [17]. All the alcoholic patients were clinically normal with no past history of psychiatric disorders. The clinical assessment included detailed medical history, neurological, and neuropsychological examinations, and laboratory tests (routine hematology and biochemistry screen, thyroid function tests). All subjects underwent a thorough medical, psychiatric, and neurological examination by a senior psychiatrist. Exclusion criteria for all subjects included any kind of neurological symptoms (cerebellar, sensory, or motor dysfunctions), history of psychiatric disorder (other than alcohol dependence for patients), medical conditions that may alter cerebral function (i.e., cardiovascular, endocrinological, autoimmune, or oncological diseases), and brain trauma (seizures, degenerative disease, previous head injury with loss of consciousness). In addition to these examinations, the presence of visible abnormalities on T2 images (assessed by a radiologist), pacemaker, bypass surgery or metallic implants that would preclude MRI
Neuroradiology (2014) 56:69–77
71
Table 1 Characteristics of study groups Characteristics
Alcoholdependent subjects (n =18)
Healthy subjects (n =18)
Age(years) Body mass index (kg/m2) Education (years) AUDIT Alcohol consumptiona Duration of dependence (years) Abstinence (weeks) Age (years) at first drinking Age (years) at the onset of dependence Biological variables: g-Glutamyl-transferase Alanine aminotransferase (U/l) Aspartate aminotransferase (U/l) Aspartate aminotransferase/alanine aminotransferase
36.5±5 24.5±4.1 10±1.89 30.2±4.6 153.3±19.2 4.43±1.3
35.2±3.7 24.6±3.5 10±1.85 2±1.3b,c
Laboratory norms
17.47±4.39 23.7±3.1 33.2±6.2
51.3±2.1 26.4±14.1
≤53 ≤38
27.3±16.1
≤40
1.03±0.27
≤2
a
Consumption was defined as grams of pure alcohol per day during 3 months preceding detoxification
b p value (≤0.05) for between-group comparisons performed using twosample t test c
Average score of five control subjects who were occasional drinkers
scan or substance abuse (other than alcohol and tobacco) resulted to the exclusion of the subject. Therefore, the set of exclusion criteria led up to select alcohol-dependent patients without apparent neurological conditions. The local ethics committee approved the study and written informed consent was obtained from all participants after the procedures had been fully explained. Abstract reasoning task The abstract reasoning task was divided into two conditions: reasoning condition (R) and control condition (C). Each condition consisted of two phases: stimulus and response. Subjects were presented with the stimulus phase for 4 s, immediately followed by the response phase for 3.5 s. During the stimulus phase, subjects saw three pictures simultaneously in one row across the top of the screen. During the response phase, subjects saw two possible answer choices presented simultaneously in one row across the bottom of the screen. To select an answer, all subjects responded with a response grip in each of their hands. They pressed the button on the left to select the answer on the left, and the button on the right to select the answer on the right. During the trials, the phrase
“NEXT?” appeared at the top of the screen for the duration of the trial. During R trials, subjects saw three sequential stimuli presented in one row at the top of the screen. During C trials, subjects saw the same stimuli in all three positions of the top row. Subjects looked at the three stimuli and determined what the fourth pictures in the sequence would be (Fig. 1). During the response period, the top row disappeared, and subjects saw two possible answer choices on the bottom of the screen. The stimuli in the R condition consisted of line drawings of shapes such as circles, squares, and triangles. Sequences were designed such that they changed along only one dimension: number, position, shading, or size [18, 19]. Trials were designed such that there were an equal number of trial types that varied across each of the four dimensions. The answers for the R trials consisted of the correct response, and a foil that was either a picture in the presented sequence, or a variation along another dimension. C stimuli were created by randomly selecting one of the stimuli from the R sequences. As with the R trials, the answers for the C trials consisted of the correct response and a foil [19]. The control task was designed to require similar processing in terms of visual encoding, decision process, and motor response execution, without the need for reasoning. All subjects were first introduced to the tasks and response device in out-of-magnet training that included introduction and instruction slides as well as sample trials of Reasoning and Control trials on a laptop to make sure that they understood the task. Training was geared at thoroughly familiarizing participants with the modes of stimulus presentation, task requirements, and response options.
Scanning protocol The participants were scanned inside a 3-T whole-body MRI system (Magnetom Skyra, Siemens, Germany) equipped with a circularly polarized 20 channel matrix head and neck coil and 45 mT/m actively shielded gradient system. Subjects lay in the supine position with their heads supported and immobilized within the head coil using foam-pads (vendor provided), to minimize head movement and gradient noise. Thirty-six axial slices parallel to the bicommissural plane through the frontoparietal cortex covering the whole brain volume using gradient echo-based interleaved EPI sequence (matrix=64×64, field of view=210 mm, TE=36 ms, TR=3 s, flip angle=90°, slice thickness=3 mm, voxel size=3.28×3.28×3 mm3) were obtained. For anatomical reference, a T1-weighted 3D gradient echo sequence (MPRAGE: Magnetization Prepared Rapid Acquisition Gradient Echo, 160 sagittal slices, slice thickness= 1 mm, field of view=256 mm, TR=1900 ms, TE=2.07 ms) image data set was acquired coplanar with the functional scan, to allow for spatial registration of each subject’s data into a standard coordinate space.
72
Neuroradiology (2014) 56:69–77
Fig. 1 Experimental task. For all stimuli, subjects saw an instructional cue word and three pictures at the top of the screen. After 4 s, the top row disappeared, and two answer choices appeared at the bottom of the
screen. For both reasoning (R) and control (C) trials, the word “NEXT?” cued participants to determine the next picture in the sequence
fMRI Protocol
Data analysis
Block paradigm (BABABABABABAB) with alternating six phases of activation (A) and baseline (B) was chosen. 190 sequential image volumes (belonging to six alternating cycles+one baseline for eliminating T1 saturation effects and acclimatization of the patient to the gradient noise) were taken. The baseline (30 s) and the experimental task (60 s) blocks consisted of four and eight stimuli each. The total fMRI acquisition time was 9 min and 34 s including the activation (6 min) and baseline phases (3.5 min) along with an introductory screen (4 s) at the beginning of task. Stimuli were presented using fMRI hardware from NordicNeuroLab and the subject’s response was monitored with the help of Nordic response device system (NordicNeuroLab (http://www.nordicneurolab.com/Productsand Solutions/Nordic fMRI solution/index.aspx)). Timing of stimuli presentation was synchronized with the scanner image volume acquisition rate. Response times (RT) and response accuracies (RA) were recorded and stored on a PC outside the magnetic resonance scanner room, for off-line statistical analysis.
At first, fMRI data were transformed into NIFTI format. fMRI image data were processed with Statistical Parametric Mapping (SPM8, Wellcome Department of Cognitive Neurology, London, UK, http://www.fil.ion.ucl.ac.uk/spm) software package implemented in MATLAB R2008a (Version 7.6.0, Math works, Sherbon, MA). The first ten brain volumes (one baseline) of each fMRI data set were discarded to remove the initial transit signal fluctuations and subsequent images were realigned within the session to remove any minor movements. Translational or rotational movement of ±1.5 mm was not considered for analysis. The T1-weighted high-resolution anatomical images were co-registered with fMRI images and spatially normalized according to the Montreal Neurological Institute (MNI) brain template. The time-course images were normalized using the same normalization parameters and then smoothed with a 6×6×6 mm3 (full width at half maximum) Gaussian smoothing kernel. The EPI-images were high-pass filtered (128 s) to remove artifacts due to cardio respiratory and other cyclical influences. A statistic parametric map
Neuroradiology (2014) 56:69–77
(SPM) was generated for each subject under each condition by fitting the stimulation paradigm to the functional data, convolved with a hemodynamic response function. Conditionspecific effects at each voxel were estimated using the general linear model [20]. Individual first-level contrast images were generated for the abstract reasoning task versus baseline contrast (FWE-corrected, p 10 voxels for two-sample t test, which resulted in a corrected threshold of p alcohol dependents When subtracting the alcohol group from the control group, there remained no significant areas showing greater activation.
Results Drinking history variables for the alcohol-dependent subjects are presented in Table 1. The alcohol-dependent and control groups were similar with respect to mean age, years of education, and body mass index. Behavioral performance During the active phase (abstract reasoning trials), there was no difference with respect to accuracy (two-sample t test, p ≤0.12) between alcohol-dependent subjects (percentage accuracy: mean=63.52, SD=3.37; total no. of responses: mean=41.23,
Fig. 2 3D-rendered whole brain activation map for alcohol-dependent subjects. Activation maps are displayed at a threshold of t >10, p 10, p
