Callosal Contributions to Simultaneous Bimanual Finger Movements

Subjects.

Fourteen patients (nine females and five males; age range, 28–51 years; mean ± SD, 37.4 ± 5.8 years; mean ± SD disease duration, 99.4 ± 55.8 months) affected by MS and 10 healthy volunteers (eight females and two males; age range, 26–47 years; mean ± SD, 33.4 ± 6.0 years) were included in this study. All subjects were naive to the specific purpose of this study. Informed consent was obtained according to our institution policy and to the Declaration of Helsinki. Inclusion criteria were the following: both sexes, age older than 18 years, and right-handed according to the Edinburgh Handness Inventory (Oldfield, 1971), and patients had definite MS as determined by Poser's criteria (Poser et al., 1983), with stable phase of the disease without relapses or modifications in neurological examination in the last 3 months. Exclusion criteria comprised the following: magnetic resonance imaging (MRI) contraindications (claustrophobia, pregnancy, and presence of a pacemaker and other metallic parts not magnetic resonance compatible), inability to comfortably perform simple finger-tapping movements at a frequency of ∼1 Hz, and use of medications known to influence cerebellar function and/or muscle tone, such as antiepileptic drugs, benzodiazepine, antidepressants, Beta blockers, drugs for spasticity. All patients were clinically evaluated at the time of enrolment with the Kurtzke expanded disability status scale (Kurtzke, 1983). Clinical disability of MS patients ranged from mild to moderate as reported in Table 1.

Apparatus.

Subjects wore a sensor-engineered glove (patent pending, number TO2005A00368, 31/05/2005) on both hands. Data were acquired at 1 kHz (board 800008B-01; National Instruments, Austin, TX); custom-made software generated pacing signals and recorded the time of each tone and the time of each finger touch (Bove et al., 2007).

Behavioral paradigm.

Subjects were asked to perform repetitive finger opposition movements of thumb to index, medium, ring, and little fingers with the two hands simultaneously and to pace the movements with a metronome tone delivered through isolation headphones. Three different rates were randomly presented (1, 1.5, and 2 Hz); the choice of this range of rates was based on the experimental evidence that MS patients were not fully compliant with movements faster than 2 Hz. In fact, we were mainly interested in studying the motor behavior during the performance of correct sequences in which the percentage of number of errors was limited but meaningful as comparison parameter. Although these rates do not differ much from one another, it was shown that, from 1 to 2 Hz, normal subjects began to significantly change their motor behavior during the execution of unimanual repetitive finger movements (Bove et al., 2007).

Subjects were instructed through a video that showed a hand executing the task three times, and then they practiced the task at their own pace. Training ended generally within 2 min, when they were able to execute the task without errors. Then, they were seated in a comfortable chair in a quiet and darkened room. An eyes-closed paradigm was chosen to exclude possible confounding effects attributable to the integration of acoustic and visual information and to prevent MS patients from compensating for possible motor impairments by visual inspection. Testing session included three 45 s trials (one per rate). At the beginning of each block, three tones were given to signal the task pacing; subjects were instructed to listen to and to wait for these three tones and to start pacing their first movement with the fourth tone. A signal indicated the end of each block; blocks were separated by 2 min rest. Subjects were also instructed to stop a sequence and to quickly restart from the first finger when they made errors such as skip of a finger, a double touch to the same finger, and a touch of the thumb to two or more fingers at once.

Motor behavioral data analysis.

To describe the different aspects of the studied motor task, data were processed with customized software, and a number of parameters were defined for the specific analysis. Bimanual coordination was studied by means of the interhand interval (IHI) to assess possible impairments already observed in patients with damages at the level of corpus callosum (Eliassen et al., 2000). For bimanual coordination, we meant when two hands are able to perform a bimanual motor task with a high level of synchronization; specifically in our case, it means synchronized movements of equivalent fingers in the two hands. Thus, the absolute value of the timing difference between these movements was chosen to demonstrate that the two hands do not make the same movement exactly in a simultaneous way. According to this definition, the larger is the IHI absolute value, the more severe is the impairment in bimanual coordination. Particularly, we identified two measures of IHI: IHI_onset, defined as the absolute time difference between the touch onset occurring in the left hand and the corresponding touch in the right hand, and IHI_offset, defined as the absolute time difference between the finger touch offset occurring in the two hands.

Finally, to investigate the presence of possible motor impairments of MS patients in the performance of finger opposition movement sequences, we studied both the finger-touching phase and the transition phase from a finger to the successive one in the sequence. These two phases correspond, respectively, to the touch duration (TD), computed as the contact time between thumb and another finger, and the intertapping interval (ITI), the time between the end of the contact of thumb and another finger and the beginning of the successive contact. Furthermore, temporal and spatial accuracy was investigated by evaluating the relative timing and the errors made during the motor sequence performance. Relative timing was defined as the time between the touch onset and the corresponding acoustic cue: it was negative when the touch preceded the metronome tone and positive when the touch followed the metronome tone. For each trial, we computed the percentage of errors over the total number of touches made during the motor task, and the incorrect sequences were discarded from additional analysis.

Magnetic resonance imaging protocol.

MRI was performed on a 1.5 tesla MR system (Signa Excite; GE Medical Systems, Milwaukee, WI). Each subject underwent brain MRI examination including coronal T₁-weighted three-dimensional spoiled gradient recalled echo imaging [slice thickness, 1.5 mm; repetition time (TR), 28 ms; echo time (TE), 6 ms; flip angle, 20°; field of view (FOV), 260 mm; matrix, 256 × 256], axial proton density/T₂-weighted imaging (slice thickness, 3 mm; TR, 3000 ms; TE, 16.1–96.8 ms; flip angle, 90°; FOV, 250 mm; matrix, 256 × 256), and, in patients only, T₁-weighted images after gadolinium chelate administration at the dosage of 0.1 mmol/kg. Diffusion tensor imaging was performed by using single-shot spin-echo echoplanar sequences, with diffusion gradients applied in 15 noncollinear directions (b = 1000 s/mm²) and two baseline acquisitions without diffusion gradients (slice thickness, 2 mm; TR, 16,000 ms; TE, 105 ms; flip angle, 90°; FOV, 240 mm; matrix, 256 × 256). All the series covered the whole brain and were acquired without gap between slices.

Magnetic resonance DTI analysis.

Data obtained with diffusion tensor imaging were processed by using FDT, a software tool for analysis of diffusion-weighted images, part of FSL (Smith et al., 2004). High-diffusion anisotropy is primarily expected in the CC (Chepuri et al., 2002), because most axonal fibers are directed horizontally. Furthermore, the degree of diffusion anisotropy increases with normal myelination processes and with organization of axons during development (Neil et al., 1998), and CC is a highly organized and myelinated structure (Aboitiz et al., 1992). For these reasons, FA is particularly indicated to study pathological processes involving the CC; a reduction of FA in the CC of MS patients has been shown previously (Ge et al., 2004; Oh et al., 2004; Hasan et al., 2005). Thus, we obtained FA parametric maps to quantify damage along the corpus callosum.

Subregions of the CC were selected similarly to a previously described method that used a geometrical scheme to divide the CC into five parts on the basis of an “anterior-to-posterior” distance, measured on a curve from the tip of the rostrum to the tip of the splenium, with the first subregion accommodating the rostrum and the entire genu (de Lacoste et al., 1985). Regions of interest (ROIs) were drawn in the CC of each subject on the midsagittal plane of FA maps (Fig. 1A), excluding voxels at the edge to minimize possible partial-volume artifacts; the anatomic correspondence of the ROIs was visually checked on the T₂-weighted baseline echoplanar images. One ROI (CC1) included the genu and the rostrum, another ROI (CC5) covered the splenium, and three ROIs (CC2, CC3, and CC4) were identified in the remaining portion of the CC, mainly represented by the body, subdividing the anterior-to-posterior distance into three parts.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Representative fractional anisotropy map of a control subject with the subdivision of corpus callosum in five regions of interest (CC1–CC5) (A) and paths connecting callosal regions with cortical areas obtained by probabilistic tractography and displayed in axial, coronal, and sagittal plane (B).

Moreover, for control subjects, we complemented the anatomic subdivision of the CC with information derived from tractography, to reveal paths connecting callosal regions with cortical areas (Fig. 1B). To achieve this, each single ROI was used as seed mask for probabilistic tractography (Behrens et al., 2003). Tractography identifies white matter bundles mostly in agreement with classical anatomical knowledge even when technical limitations might lead to false-positive or negative results attributable to noise, voxel contamination effects, and coexistence within a voxel of multiple fibers with complex architecture. Mean FA values were then calculated in the whole CC and within each ROI for all the enrolled subjects.

Statistics.

The changes in IHI_onset, IHI_offset, TD, ITI, relative timing, and the percentage of errors made during the bimanual motor task were subjected, separately, to ANOVA. Precisely, bimanual coordination parameters were separately analyzed by means of a one-way ANOVA with the variable group (MS patients and control subjects) used as between-subject factor and metronome rate (1, 1.5, and 2 Hz) as within-subject factor. Furthermore, the IHI_onset and IHI_offset were compared between them by means of a two-way ANOVA with group (MS patients and control subjects) as between-subject factor and metronome rate (1, 1.5, and 2 Hz) and IHI (IHI_onset, IHI_offset) as within-subject factors. For single-hand kinematics parameters (TD, ITI, relative timing, and percentage of errors), a two-way ANOVA was performed, separately for each parameter, with group (MS patients and control subjects) as between-subject factor and metronome rate (1, 1.5, and 2 Hz) and hand (dominant and nondominant hand) as within-subject factors. When ANOVA gave a significant result (p < 0.05), post hoc Newman–Keuls test was used to assess significant differences.

After assessing that FA values did not show a normal distribution by using the Shapiro–Wilk normality test, the comparisons to find out differences in the mean FA values both in the whole CC and within the single CC ROIs between MS patients and control subjects were performed by means of the nonparametric Mann–Whitney U test. Then, Pearson's correlation coefficients were calculated between the FA in each ROI of the CC and the movement parameters for MS patients. Correlations were performed for each metronome rate; in addition, when the statistical analysis indicated that a specific motor behavior parameter was not significantly influenced by rate, an average on the values obtained for the three metronome rates was performed and then correlated with FA values of the CC ROIs.

Numerical data are mean ± SE unless otherwise stated.