Reduced interhemispheric inhibition in mild cognitive impairment
In mild cognitive impairment (MCI), the corpus callosum is known to be affected structurally. We evaluated callosal function by interhemispheric inhibition (IHI) using transcranial magnetic stimulation (TMS) in MCI patients. We investigated 12 amnestic MCI patients and 16 healthy age-matched control subjects. The IHI was studied with a paired-pulse TMS technique. The conditioning TMS was given over the right primary motor cortex (M1) and the test TMS over the left M1. Motor evoked potentials were recorded from the relaxed first dorsal interosseous muscle. We also studied other motor cortical circuit functions; short-latency afferent inhibition (SAI), short-interval intracortical inhibition (SICI) and intracortical facilitation (ICF). Both the amount of IHI and SAI were significantly reduced in MCI patients as compared with control subjects, whereas SICI or ICF did not differ between them. The degree of IHI significantly correlated with neither the mini-mental state examination score nor the degree of SAI. Our results suggest that transcallosal connection between bilateral M1 is primarily involved in MCI, regardless of SAI dysfunction.
KeywordsMild cognitive impairment Alzheimer’s disease Corpus callosum Interhemispheric inhibition Short-latency afferent inhibition Transcranial magnetic stimulation
Mild cognitive impairment (MCI) is a cognitive disorder, which is considered to be a transitional state between normal aging and dementia such as Alzheimer’s disease (AD) (Winblad et al. 2004; Gauthier et al. 2006). Recent studies using new magnetic resonance imaging (MRI) techniques, such as voxel-based morphometry and diffusion tensor imaging studies, revealed specific changes in the white matter including the corpus callosum in MCI and AD patients (Di Paola et al. 2010). Pathological study also showed corpus callosum atrophy in AD, which must cause interhemispheric disconnection (Tomimoto et al. 2004). However, it is still unclear what functional changes reflect the corpus callosal atrophy in MCI and AD. We hypothesize that in these disorders, functions of cortico-cortical connections are damaged. In this paper, we evaluated the cortico-cortical connections in MCI using transcranial magnetic stimulation (TMS).
TMS, a non-invasive human brain stimulation method, is applied to AD patients to evaluate cortical excitability (Pepin et al. 1999; Alagona et al. 2001). Especially, short-latency afferent inhibition (SAI) (Tokimura et al. 2000) is known to be reduced in AD (Di Lazzaro et al. 2002). However, interhemispheric inhibition (IHI) between bilateral primary motor cortices (M1s) has not been studied. In this paired-pulse TMS technique, motor evoked potentials (MEPs) to a test TMS over one hemisphere are suppressed by a conditioning TMS over the other hemisphere at interstimulus intervals (ISIs) around 10 ms (Ferbert et al. 1992). The inhibition is considered to be produced by an intracortical inhibition at the target M1 activated by the excitatory transcallosal inputs from the conditioning M1 (Chen 2004).
In the present paper, we evaluated callosal function in MCI patients using IHI. We also compared other parameters such as SAI for estimating cortico-cortical connection between the sensory cortex and M1, and short-interval intracortical inhibition (SICI) and intracortical facilitation (ICF) for estimating intracortical circuit functions of M1 (Kujirai et al. 1993).
We examined 12 amnestic MCI patients and 16 healthy elderly control subjects. Subjects were all right-handed by self-report. MCI patients were recruited from the departments of Neurology, the University of Tokyo and the Osaka City University hospitals. The diagnosis of amnestic MCI was made by board-certified neurologists on the basis of the consensus recommendation by the International Working Group on Mild Cognitive Impairment (Winblad et al. 2004). The mean age was 72.3 ± 9.3 years (mean ± SD) for MCI patients (7 women and 5 men) and 68.1 ± 4.9 years for control subjects (11 women and 5 men). All patients complained with amnesia, and the average mini-mental state examination (MMSE) (Folstein et al. 1975) score at the time of examination was 25.3 ± 2.4 (range 21–29). The average duration of the disease was 3.3 ± 1.8 years (range 1 to 6). In all the patients, MRIs revealed hippocampal atrophy. It was judged by board-certified neuroradiologists, who had no clinical information, by visual inspection. Some other neuroimaging studies confirmed their findings. Among 12 MCI patients, all 7 patients who were studied with Pittsburgh Compound-B (PIB) positron emission tomography (PET) had abnormal amyloid deposition (Klunk et al. 2004). The other 5 patients underwent 123I-IMP single-photon emission computed tomography (SPECT), and a perfusion reduction at the temporo-parietal cortex was revealed in all of them. Based on these results, we concluded that all our patients were high- to intermediate-grade MCI due to AD pathology following the recommendations from the National Institute on Aging and Alzheimer’s Association workgroup (Albert et al. 2011). Among 12 MCI patients, 5 patients had taken donepezil regularly.
All participants or their caregivers gave their written informed consent to participate in this study. The procedures done here were approved by the Institutional Review Boards of both Universities in accordance with the ethical standards of the Declaration of Helsinki on the use of human subjects in experiments.
Surface electromyograms were recorded from the bilateral first dorsal interosseous (FDI) muscles with 9-mm-diameter Ag/AgCl surface electrodes placed with a belly-tendon montage. Responses were input to an amplifier (Biotop; GE Marquette Medical Systems Japan, Japan) through filters set at 100 Hz and 3 kHz. They were then digitized with a sampling rate of 10 kHz and stored in a computer for later offline analyses (TMS bistim tester; Medical Try System, Japan).
Transcranial magnetic stimulation
Throughout the experiments described below, subjects were seated on a comfortable chair and the FDI muscles were relaxed, as confirmed by an oscilloscope monitor. For TMS stimulation, monophasic TMS pulses were delivered by magnetic stimulators (Magstim 200; Magstim Co., Whitland, Dyfed, UK). The intervals between the trials were set at 8 ± 0.5 s. In advance, we measured the resting and active motor thresholds (RMT and AMT) and the central motor conduction time (CMCT) for each muscle to exclude cortico-spinal tract impairments. RMT was determined as the lowest stimulator output intensity capable of eliciting MEPs of 50 μV peak-to-peak amplitude in the relaxed FDI muscle in more than 5 of 10 consecutive trials. AMT was determined as the lowest stimulator output intensity to evoke MEPs of 100 μV peak-to-peak amplitude when the participant maintained a very slight contraction of FDI muscle (5–10% of the maximum voluntary contraction) in more than 5 of 10 consecutive trials. We also evaluated the CMCT as described previously (Ugawa et al. 1989).
Experiment 1: Interhemispheric inhibition (IHI)
The test stimulus (TS) was given over the left M1, and the conditioning stimulus (CS) over the right M1 preceding TS by 4, 6, 8, 10 and 12 ms. For both stimuli, we used a figure-of-eight coil (outer diameter of each wing was 7 cm) positioned over the optimum point for FDI (about 5 cm lateral to the vertex). The coil for TS was placed tangentially over the scalp and angled 45° to the parasagittal plane so that current flowed in an anteromedial-to-posterolateral direction at the center of the coil. The coil for CS was set toward the sagittal plane so that current flowed in a medial-to-lateral direction at the center of the coil. The intensities of TS and CS were both adjusted to elicit MEPs of 0.5–1 mV peak-to-peak amplitude in the relaxed muscles on average when they were given alone. Eight conditioned trials for each ISIs were randomly intermixed with 16 unconditioned trials in which TS was delivered alone.
Experiment 2: Short-latency afferent inhibition (SAI)
We compared the amount of IHI with SAI, which is previously known to be reduced in AD (Di Lazzaro et al. 2002), and analyzed whether or not the two measures of cortical inhibition were correlated with each other. For SAI, TS was TMS over the left M1, and CS was the right median nerve stimulation at the wrist. The median nerve was stimulated with a 0.2-ms-duration square-wave electric pulses (cathode proximal) preceding TS by 20 ms. The CS intensity was set at just above the level to evoke a visible thumb twitch. Ten conditioned trials were randomly intermixed with 20 unconditioned trials in which TS was delivered alone.
Experiment 3: Short-interval intracortical inhibition (SICI) and intracortical facilitation (ICF)
We evaluated whether intracortical excitability of M1 was also abnormal or not. Nine out of 12 MCI patients and 13 out of 16 control subjects participated in SICI and ICF experiments. Paired pulses of CS and TS were given through the same coil over the left M1 using a bistim module (Magstim Co., Whitland, Dyfed, UK). We set the CS at 90% of AMT. CS was given before the TS at ISIs of 2, 3 and 4 ms for SICI, and 8, 10 and 15 ms for ICF. Nine conditioned trials for each ISIs were randomly intermixed with 18 unconditioned trials in which TS was delivered alone.
Age, body height, RMT, AMT, CMCT and control MEP size for each experiment were compared between MCI and control groups using Student’s t test. Among MCI groups, age, MMSE score and disease duration were compared between patients with and without donepezil using Student’s t test.
For the three experiments, the ratio of the mean peak-to-peak amplitude of conditioned MEPs to that of unconditioned MEPs was calculated for each ISI in each subject. We compared MCI and control groups using this grand mean of MEP size ratio for each ISI in every experiment.
Experiments 1 and 3
To compare the conditioning stimulus effects on the MEP size ratio between MCI patients and control subjects, we used two-way repeated measures analysis of variance (ANOVA) using group (MCI patients and control subjects) as a between-subjects factor and ISI as a within-subject factor. The dependent variable was the MEP size ratio. When necessary, Greenhouse–Geisser correction was used to correct for non-sphericity.
To compare the SAI effect between MCI patients and control subjects, we used Student’s t test. The size ratios were also compared between patients with and without donepezil.
Correlations between IHI and MMSE or SAI were analyzed using linear regression analyses. IHI was represented by the size ratio at an ISI of 10 ms in this analysis.
Statistical analyses were performed using PASW Statistics 18.0.0 (IBM Corporation, NY, USA). P values less than 0.05 were judged as significant.
Comparison of the electrophysiological values between the groups
MCI (n = 12)
Control (n = 16)
72.3 ± 9.3
68.1 ± 4.9
Body height (cm)
157.9 ± 6.4
156.5 ± 7.6
46.3 ± 12.5
46.9 ± 7.7
32.0 ± 7.9
33.6 ± 6.6
6.7 ± 1.3
6.8 ± 0.49
MEP size (mV)
Experiment 1 TS
0.87 ± 0.61
0.75 ± 0.35
Experiment 1 CS
0.53 ± 0.27
0.63 ± 0.38
Experiment 2 TS
0.87 ± 0.41
0.77 ± 0.36
Experiment 3 TS
0.68 ± 0.35
0.55 ± 0.39
Characteristics of mild cognitive impairment patients
With donepezil (n = 5)
Without donepezil (n = 7)
76.2 ± 3.7
69.4 ± 11.3
25.0 ± 2.2
25.4 ± 2.6
Disease duration (years)
4.2 ± 1.5
2.6 ± 1.7
Experiment 1: IHI
Experiment 2: SAI
Experiment 3: SICI and ICF
In the present study, we first showed that IHI was significantly reduced in MCI patients. SAI, another cortico-cortical connection parameter between sensory cortex and M1, was also reduced in MCI patients. However, their degrees of reduction did not correlate with each other. In contrast, SICI and ICF, reflecting the motor cortical intracortical circuit function, were not affected in MCI. We, thereafter, will discuss the above findings separately.
Abnormal IHI in MCI
Since IHI reflects the transcallosal pathways function, the above result may indicate a damage of interhemispheric cortico-cortical connection of M1 in MCI. Which mechanisms are responsible for this abnormality?
The first possibility is that the corpus callosum itself is damaged in MCI and that IHI was reduced due to reduced inputs from the contralateral M1. Many MRI studies showed structural changes in the corpus callosum in MCI and AD (Chua et al. 2008; Di Paola et al. 2010; Douaud et al. 2011). Other functional MRI study showed effective connectivity is altered between primary sensorimotor cortices in amnestic MCI patients (Agosta et al. 2010). These reports support our hypothesis that the corpus callosum may be primarily involved in MCI.
The second possible explanation of the present finding is that some interneurons connecting callosal fibers in M1 are damaged. Those interneurons may be directly involved or functionally involved secondary to other interneuronal dysfunction. In a triple-pulse TMS study, for example, the interneurons related with SICI influenced some functions of the interneurons related to IHI (Lee et al. 2007). In our subjects, SICI tended to be reduced, but this reduction was not statistically significant. From these results, we cannot completely exclude a mild SICI involvement because the number of the subjects may be not many enough for detecting slight SICI abnormality. However, they were not strongly affected, and their dysfunction alone should not explain abnormal IHI shown here. Based on these arguments, we conclude this possibility unlikely even though this may explain our finding only partly.
Third, we also consider the possibility that reduced SAI could affect the amount of IHI. However, it is unlikely because the degree of abnormality did not significantly correlate between these two evaluations. We think that reduced IHI is independent of the SAI reduction. The SAI should relate to some cortical cholinergic circuit because donepezil normalized SAI in AD patients (Di Lazzaro et al. 2002, 2004, 2005; Nardone et al. 2008). The reduced IHI must not relate to abnormal cholinergic circuit. In addition, in our patients, neither IHI nor SAI was influenced by donepezil treatment.
No correlation was found between the amount of IHI and MMSE score in this study. This may be because MMSE score is not able to detect small difference in mental function in MCI. This suggests that IHI could be used for early detection of AD pathology that is not able to be picked up by ordinary mental tests.
Other cortical excitability parameters using TMS
In our MCI patients, both SICI and ICF were normal. Some previous papers reported a reduction in SICI in early-onset AD (Pierantozzi et al. 2004) and MCI converted to AD (Olazarán et al. 2010) even though there was high inter-individual variability. No consistent result has been reported concerning the ICF. They concluded that SICI and ICF were not good tools for the diagnosis of early-stage AD (Olazarán et al. 2010). In our patients, both SICI and ICF were normal; though SICI had a tendency to be reduced in MCI, SICI might be mildly reduced in MCI. We consider, however, that neither the intracortical circuits for SICI nor those for ICF are definitely involved in MCI or early stage AD.
On the other hand, SAI is reduced in AD patients, whereas a previous paper showed normal SAI in MCI patients (Sakuma et al. 2007). This difference is probably because of the heterogeneity of MCI pathogenesis. Since more than half of our patients showed evidence of amyloid deposition, we suppose that the converting rate to AD is higher in our group. SAI may be abnormal even at an early stage in MCI patients due to AD pathology.
Functions of cortico-cortical connections might be more damaged than intracortical functions in MCI. The combination of abnormalities of the two cortico-cortical functions, SAI and IHI, may be useful for the diagnosis of early-stage MCI converting to AD. Based on the above arguments, we conclude that transcallosal connection between bilateral M1 is primarily involved in MCI, regardless of SAI dysfunction.
Part of this work was supported by the following: Research Project Grants-in-aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan (No. 22390181, No. 22590954, No. 20591019); the Research Committee on rTMS Treatment of Parkinson disease from the Ministry of Health, Labour and Welfare of Japan (H20-023); the Research Committee on Dystonia from the Ministry of Health, Labour and Welfare of Japan; the Research Committee on Intractable Pain from the Ministry of Health, Labour and Welfare of Japan; the Research Committee on Degenerative Ataxia from the Ministry of Health and Welfare of Japan; and the Global COE Program (Comprehensive Center of Education and Research for Chemical Biology of Diseases) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.
Conflict of interest
There is no conflict of interest.