Repetitive Transcranial Magnetic Stimulation in Initial Alzheimer S Disease

Cortical excitability in very mild Alzheimer’s disease: a long-term follow-up study

J. Olazarána ,b , J. Prieto cb, I. Cruza, A. Esteban cb

Services of aNeurology and cbClinical Neurophysiology, Gregorio Marañón General University Hospital; and bNeurology Clinic, Hermanos Sangro Specialties Center; , Madrid, Spain

Corresponding author: Javier Olazarán Servicio de Neurología HGU Gregorio Marañón Dr. Esquerdo 46 28007 Madrid [email protected]

Tel.: 00-34-91-7306051 Fax: 00-34-91-4336033

WORD COUNT: 2,813

RUNNING TITLE: Cortical excitability in Alzheimer’s disease

ACKNOWLEDGEMENT. This study was supported by Pfizer and EisaiInc.

1 ABSTRACT

Background: Measurement of Cortical motor cortex excitability , measured byusing paired-pulse transcranial magnetic stimulation (pTMS), has been proposed for the early diagnosis of Alzheimer’s disease (AD) and could also be useful forto monitoring treatment response and disease progression. However, studies conducted at the pre- dementia stage of AD are scarce, very few long-term data are available, and correlations between cortical excitability and cognitive performance have not been addressed yet. Methods: Eleven patients with mild cognitive impairment (MCI) that converted to AD-related dementia and 12 elderly people were selected for this study. Cognitive assessments and pTMS were conducted at baseline in the two groups and also after 4four and 21 months of donepezil treatment with donepezil in the AD group. Non- parametric statistics were used to compare cortical excitability between the two study groups at baseline and to analyse disease courseevolution in the AD group. Correlation analysis was performed to investigate associations between cortical excitability and cognitive performance. Results: Short-interval intracortical inhibition (SICI) and intracortical facilitation (ICF) were reduced in AD patients. However, there was high inter-individual variability, and statistical significance was only attained at a 2- ms inter-stimulus interval (ISI). A trend towards recovery of 2 -ms SICI after donepezil treatment was observed after treatment with doenepezil. Baseline cortical excitability at 300 ms was associated with better cognitive performance in AD patients. Conclusions: Although the present results do not support a role offor pTMS in the early diagnosis of late-onset AD, a potential role of pTMS to both in prediction of treatment response and understanding of disease mechanisms of disease emerged.

KEY WORDS: Cortical excitability, Alzheimer’s disease, paired-pulse transcranial magnetic stimulation, mild cognitive impairment

2 INTRODUCTION Motor cortex excitability, which can be evaluated by means of transcranial magnetic stimulation (TMS), is usually increased in Alzheimer’s disease (AD), possibly as a consequence of neuronal dysfunction at different cortical and spinal levels [1-3]. Paired-pulse TMS (pTMS) makes it possible to study permits an estimation of intracortical GABAgabaergic and glutamatergic neural mechanisms [4-6]. Using Tthis technique, a reduced intracortical inhibition (ICI) was found that correlated with the severity of AD-related dementia is altered in AD [7]. More recently, pTMS and has been proposed as an ancillary tool in the early diagnosis of AD [8]. However, research is was limited to short interstimulus intervals (ISIs) and the reported results weare controversial [8-10]. Moreover, the cognitive correlates of the excitability curve to pTMS excitability curve have were not been analyzsed, and studies exclusively conducted in pre-dementia AD are lacking. Mild cognitive impairment (MCI) is a transitional clinical state between normal aging and dementia [11]. Since more than 50% of patients with MCI will develop dementia -mainly due to AD,- the search ofor clinical and paraclinical markers in of thoese patients is the subject constitutes a matter of intense research [12]. Cholinergic dysfunction is the earliest biochemical indicatorhallmark of AD, and cholinergic compounds are the only drugs that have demonstrated consistent benefits at the mild dementia stage of disease [13]. However, consistent benefits of cholinesterase inhibitors (ChEI) in MCI werehave not demonstrated in MCI, maybe due to the etiological heterogeneity of the MCI construct [14] . In light of the existing connections between hippocampal cholinergic hippocampal pathways and GABAgabaergic interneurons, a role ofor pTMS in the prediction of both etiology and treatment response of MCI patients could be proposed for patients with MCIspeculated [15]. We The main aim of this study was to investigated a potential role for the of pTMS recovery curve into the identificationying AD of patients with AD in the at MCI clinical stage. Previously studiedFormerly investigated ISIs were expanded, short- and long-term changes in response to cholinergic treatment were analyzsed, and correlates between cognitive performance and pTMS data were explored.

3 PATIENTS AND METHODS Right-handed patients with cognitive complaints and some evidence of cognitive impairment were recruited from a Hermanos Sangro Neurology Clinic. Their se patients were at the 0.5 stage onf the Clinical Dementia Rating (CDR) Scale was 0.5 [16] and they did not have dementia according to DSM-IV-TR criteria [17]. All patients underwent Aa semi-structured interview, a neurological exam, the Mini-mental State Examination (MMSE) Test [18] and the cognitive subscale of Alzheimer’s Disease Assessment Scale (ADAS-cog) [19] were conducted in all cases. The MMSE and ADAS-cog tests are the most widely used tools to evaluate general cognition in AD. They are composed of subtests for the measurement of disturbances of temporal orientation, verbal learning, language, attention, and visuoespatial abilities. An Eelectrocardiogram, biochemical parameters (including Ca, B12, folate and TSH) and cranial computerized tomography (CT) scan were also performed, and biochemical parameters (including Ca, B12, folate and TSH) were analysed. In all these patients, AD was the suspected diagnosis in all casesmain etiological suspect. Right-handed control subjects were recruited from Gregorio Marañón Neurophysiology Service and Hermanos Sangro Neurology Clinic, and s. They were comprised patients referred due to peripheral nerve disease, tension headache, or other conditions, or were the patient’s spouses of the patients that volunteered to participatinge in this study. These control subjects did not have relevant cognitive complaints, nor did they presented any systemic or neurologic conditions that could alter cognition. As with like the AD patients, they were administered the MMSE and the ADAS-cog were applied. Medications that could alter cognitive performance or cortical excitability were not permitted in either group. Both AD patients and control subjects signed the informed consent document. Magnetic stimulation was performed using two Magstim 200 magnetic stimulators connected to a Bistim module. A circular coil was held over the vertex to stimulate the left motor cortex. Motor responses were recorded in the right first dorsal interosseous muscle. All measurements were performed at rest. Resting motor threshold (RMT) was defined as the minimum stimulus intensity that produced a motor evoked potential (MEP) of more than 50 μV in at least 50% out of six consecutive stimuli. Two magnetic stimuli were applied Tto study intracortical inhibition (ICI) and intracortical facilitation (ICF), two magnetic stimuli were given, and the effect of the first (conditioning) stimulus on the second (test) stimulus was investigated. Six

4 consecutive pairs of stimuli separated by more than 5 s from one another were applied at each of 13thirteen interstimulus intervals (ISIs) ranging from 1 to 300 ms. Signals were rectified, and average amplitudes and areas of MEP were obtained for each ISI. Two different methods were used depending on the ISI range. From 1 to 16 ms ISI, the intensity of the conditioning stimulus was 80% of RMT, the test stimulus was adjusted to 120% of RMT, and the results were expressed as a percentage of mean MEPs that were independently obtained three timesice (i.e., at the beginning, middle, and end of each pTMS study) at an intensity of 120% of RMT. From 20 to 300 ms, both conditioning and test stimuli were set at an intensity of 120% of RMT, and the results were expressed as a percentage of the response to the conditioning stimulus. A Ddecreasing inof test responses to between 0 and 5 ms ISIs represented a short-latency intracortical inhibition (SICI) and arewas attributed to the activation of intracortical GABA-A-ergic neuronal system transmission. An Iincreasinge ofin test responses to between 10 and 16 ms ISIs represented intracortical facilitation (ICF) and arewas attributed to both GABAergic and glutamatergic interaction. A Ddecreasinge ofin test responses to between 80 and 100 ms ISIs represented long-term intracortical inhibition (LICI) and arewas likely due to the GABA-B-ergic neuronal system [6]. After the baseline pTMS study, treatment with donepezil was initiated and regular follow-up assessments were conducted, including new pTMS studies four months after the initiation of donepezil and in theat long-term follow-up. After pTMS studies, regular follow-up visits or contacts were performed. Dementia was diagnosed when DSM-IV-TR criteria were fulfilled [17] and the CDR stage was ≥ 1 [16]. The NINCDS-ADRDA criteria were used for the diagnosis of AD [20]. Two-tailed non-parametric statistics were used to compare demographic, clinical, and pTMS variables in AD and control groups. Associations between age, sex, and education, and baseline neurophysiological data, were analyzsed using the Spearman correlation coefficient. Then the cognitive correlates of pTMS parameters were then systematically analyzsed at baseline in the whole sample and in AD and control groups separately. Within the AD group, longitudinal correlations between changes in pTMS parameters and changes in cognitive performance were also investigated. Given the exploratory nature of this study, multiple comparison adjustment was not performed [21]. All statistical analyses were performed using SPSS version 13.0 software (SPSS Inc, Chicago, ILllinois, USA).

5 RESULTS Seventeen patients were initially recruited. Four of them were eliminated for the following reasonsdue to: very high RMT that made the study unfeasible (one1 patient), non-AD aetiology (one1 patient), death eleven months after the baselineal study (one1 patient), or CDR stage of 0.5 at the end of follow-up contact (one1 patient). Eleven patients had developed dementia after a mean follow-up of five years (range 4 to 6) (one1 patient definite AD, 10 patients probable AD) [20]. The two remaining patients converted to a CDR stage of 0 and remainedpersisted in that stage at the last follow-up visit (3-4 years after the baselinebasal study). Therefore, these two patients were relocated to the control group, where ten control subjects had been initially recruited. The two final study groups (11 patients with very mild AD, 12 control patientsgroup) were comparable in terms of age and sex, although a trend was observed of less formal education in AD patients tended to have less formal education. As expected, cognitive performance was impaired in the AD group (Table 1). Medications wereas as follows (in parenthesis, number of patients taking drugs those medications in the very mild AD group/, control group): antihiypertensive drugs (3,/2), lipid-lowering agentshypolipemiants (3,/3), antiplatelet agentsagregants or anticoagulants (4,/1), vitamins or minerals (2,/3), selective serotonine reuptake inhibitors (2,/2), nonsteroidal anti-inflammatoriesy drugs (1,/3), antidiabetics drugs (2,/1), antiacids (2,/0), antianginal drugs (1,/1), antiprostatic agents (1/,1), antihistaminices (1,/0), hypouricemic drugs (0,/1), bisfphosfphonates (0,/1) and paracetamol (0,/1). One AD patient was a smoker. No correlations were found between age, sex, and education, and neurophysiological parameters (all p > 0.05, data not shown). RMTsesting motor thresholds were slightly decreased in AD patients (Table 1) and remained essentially unchanged during the study (all p for longitudinal intra-group comparisons > 0.10; data are not given). Reductions in SICI and ICF were observed in AD patients at baseline; however, a high inter-individual variability was observed occurred at all ISIs, and statistically significant differences or trends (p < 0.1) were only detected at 2 ms ISI (Table 2). _The Sshort-term follow-up results of pTMS and cognitive performance were conducted and analyzsed after a mean of 3.9 months (range, 3.0 to 5.3) months after baseline in 10 AD patients (one patient that who did not tolerate donepezil was not included in this analysis). Nine patients were on 10 mg/d and 1one patient was on 5 mg/d. Dato on Ccortical excitability data of for these patients, before and while on

6 donepezil, appear in Table 3 and Figure 1. With the exception of 2, 16, and 20 ms ISIs, the profile of cortical excitability of AD patients on donepezil differed from theat profile of controls more than it differed before starting donepezil. Within- group, comparisons displayedrevealed an increase in inhibition at 300 ms ISI (p = 0.028 for MEP amplitude) and an trend of increasing trende in SICI at 2 ms ISI (p = 0.059 for MEP area) (Table 3). When the patient that was on 5 mg/d of donepezile was excluded from this analysis, the results of SICI at 2 ms were reinforced (p = 0.086 for MEP amplitude), whereas all other results remained essentially unchanged. Cortical excitability could be was reassessed duringat long-term follow-up in eight AD patients (mean follow-up from baseline 20.3 months,; range, 11.6 to 27.3). Only one patient had developed dementia at that time, but cognitive performance had deteriorated in the whole group (the MMSE score had fallen from 23.0 to 21.3, p = 0.058, and ADAS-cog had worsened from 17.9 to 21.0, p = 0.150). Five patients were on donepezil 10 mg/d, one patient was on donepezil 5 mg/d, and two patients were not taking the drug. Significant changes (p < 0.05) were found for loss of SICI at 5 ms ISI, appearance of ICF at 30 ms ISI, and reduction of recovery at 300 ms ISI. When the analysis was restricted to the six patients thatwho remained on donepezil, the results remained essentially unchanged (long-term data of cortical excitability are not shown). Although markedSpite the fact that a great inter-individual variability was observed, paired TMS measurements were deemed to be reliable at the individual level. One patient that who clearly differed from the rest in terms of abscensce of SICI at 2 ms interval (185% MEP amplitude, 207% MEP area) displayed some recovery of 2- ms SICI after 4 months on donepezil (65% MEP amplitude, 67% MEP area); this that was partially lost 2 years later (98% MEP amplitude, 122% MEP area). However, these changes were not associated with a more favourable clinical course; indeed, a similar longitudinal pattern was observed in the patient thatwho was eliminated from this study due to a non-AD-related aetiology (the final diagnosis of this patient was temporal epilepsy). Remarkably, the patient whothat displayed the highest SICI at 2 ms ISI (7% MEP amplitude, 6% MEP area) was the only patient whothat did not tolerate donepezil treatment. No significant correlations were found between pTMS parameters and cognitive performance at baseline when the whole sample was analyzsed (Spearman correlation coefficients ranged from -─.293 to .338, all p values > .10; data not shown). However, when the two study groups were analyzed separately, significant correlations emerged:

7 recovery at 300 ms ISI was related to better cognitive performance in AD patients, whereas ICF from 10 to 20 ms ISIs was related to better cognitive performance in the control group (Table 4). Correlations between recovery at 300 ms ISI and cognitive performance at baseline in the AD group were especially robust (Figure 2) and remained consistent at theduring follow-up studies, particularly for the MMSE (Spearman r ranged from 0.42 to 0.72). However, neither longitudinal changes in cortical excitability at 300 ms ISI nor changes at other ISIs were consistently related to changes in cognitive performance. Of note, increased SICI increase at 2 ms ISI was not associated with an improvement in cognitive performance (correlation coefficients at 2 ms ISI ranged from -─.132 to .443, all p values > 0.10; correlations between pTMS data and cognitive evolution are not shown).

DISCUSSION pPaired-pulse TMS was used to compare the cortical excitability of AD patients at the MCI clinical stage (i.e., pre-dementia) clinical stage with that of cognitively normal elderly subjects. The profiles of cortical excitability differed between the two groups, but a high inter-individual variability was observed and statistically significant differences were reached at only 2 msec ISI only (Table 2, Figure 1). Variability in paired TMS measurements can be due to incomplete patient collaboration and examiner skills. To reduce this kind of variability, we recorded basal MEPs three timesice durialong each TMS study and calculated the averaged them. Moreover, the conditioned MEPs were also averaged from six MEPs at every ISI. Actually, tThe individual patterns of response were highly reproducible during follow-up studies;, and therefore, pTMS measurements were considered to be reliable. Nevertheless, variability of response to pTMS could also be due to a series appear as the result of diverse agents and mechanisms (e.g., individual constitution, drugs, smokingtobacco use), whichthat could influence the GABAergic and glutamatergic neuronal systems [4,22]. Although age, sex, and number of medications were similar in both our two study groups, many other potential confounders (e.g., education, kind of medication, etc.) could not be adequately controlled due to the small sample size. These confounders could have obscured differences in cortical excitability due to early AD. Two previous studies conducted paired TMS on AD patients atwith MCI and mild to moderate dementia stages. In one of those studiesthe first, no differences were found in either SICI or ICF were found when between late-onset AD patients and

8 elderlyaged subjects were compared, although but both groups displayed lower SICI than younger subjects [89]. In the secondother study, SICI was markedly reduced at 2 and 3 ms ISI in early-onset AD, and this alteration was reversed after treatment with galantamine [108]. We did not find suchas marked a reduction ofin SICI as Pierantozzi et al. [108], but we did findstill found some reduction of SICI at 2 ms ISI (Table 2). A multi-causal model in the genesis of late-onset AD (vs. compared with more genetically determined early-onset AD) and the fact that AD was confirmed during follow-up in all our patients had AD, (as opposed to compared to Pepin et al. [98],) where follow-up data are not reported, could explain the differences in the results. In agreementConsistent with Pierantozzi et al.’s results [108], we found some evidence supporting a role ofor cholinergic transmission inof SICI. A trend of SICI tended to recovery at 2 ms ISI after treatment with donepezil treatment was observed, and donepezil was not tolerated by the patient whothat displayed the highest SICI at 2 and 3 ms ISIs did not tolerate donepezile. Although neither SICI baseline values nor longitudinal changes could be related to cognitive performance, the low sample size did oes not allow us permit to definitely excluderule out a role ofor pTMS in the prediction of patient tolerability of, and even response to, to ChEI. Recently, short- latency afferent inhibition (SAI) has been proposed as a more specific TMS-related means of testing cholinergic function in AD than, compared to SICI or ICF [910,23,24]. However, two studies that comparinged SAI and pTMS in AD did not include patients at the pre-dementia stage [910,23]. Another study comparing SAI and pTMS included patients with both MCI and patients with mild dementia patients and demonstrated the superiority of SAI, butalthough 2 ms ISI was not obtained [24]. In a recent study conducted in patients with MCI of unknown aetiology, SAI was not altered in MCI [25]. MCI is a heterogeneous concept embracing cognitive decline of different causes. A considerable proportion of patients revert to normal cognition, as occurred in two of our patients, who were subsequently reallocated to the control group. For this reason, our results could be regarded as specific of MCI due to AD. Even so, a low SICI at 2 ms was demonstrated in the very mild AD group, although absence of SICI that responded to cholinergic treatment was observed in one patient who was not included in the analyses due to non-AD-related aetiology. Future research should further address the specificity of low SICI at 2 ms in patients with MCI.

9 An unexpected association between cortical excitability at 300 ms ISI and better cognitive performance was found (Table 4, Figure 2). This association was not modifiedimproved by cholinergic treatment and remained partialially during follow-up. Since this association was quite specific to the AD group, we could speculate that it was disclosingindicated damage to neuronal systems involved in cognition or, possibly, individual cognitive mechanisms to cope with cognitive losses [26]. The present study bears somehas strengths and limitations. As an original contribution, all our patients were at the pre-dementia clinical stage of AD when the baseline pTMS study was performeconducted. Moreover, AD dementia was confirmed during long-term follow-up, including autopsy in one case. The contribution of non- AD-related aetiologies in our clinically diagnosed (probable) AD patients was very unlikely. The clinical course was of typical -AD (i.e., progressive deterioration with memory deficit present at the beginning of symptoms), vascular lesions were not found in the baseline CT scan, vascular risk factors were well controlled, and vascular events were not reported at baseline or during follow-up.did not occur. As other novel contributions, tThe range of ISIs was wider compared to previous studies [7-10] ned and, to the authors’ knowledge, this wais also the first report of a correlation time that correlates between paired TMS parameters and cognitive performance were reported. Taken together, the results do not support a role ofor paired TMS in the early diagnosis of AD, althoughbut a potential role ofor paired TMS in predicting tolerability and possiblymaybe response to ChEI in late-onset AD emerged. In addition, the present unexpected findings of a correlation between cortical excitability and cognitive performance could open avenues to understanding of cognitive mechanisms of coping with disease mechanisms. This study also bearshas important limitations. Due to the small sample size and to the fact that multiple comparison adjustments were not performed, all the reported results should be confirmed, particularly those. This bears particularly true for the findings relating cortical excitability and cognitive performance. In addition, specific cognitive functions (e.g., attention), reasonably related to the GABAgabaergic and glutamatergic systems, such as attention, were not specifically addressed, and other AD- relevant features that could be related to cortical excitability (e.g., motor function and gait) were not evaluated either [27]. MoreFuture research is warranted to better understand associations between cortical excitability, cognition, and other clinical aspects of AD.

10 REFERENCES

1. De Carvalho M, de Mendonça A, Miranda PC, Garcia C, Luís ML (1997) Magnetic stimulation in Alzheimer's disease. J Neurol 244:304-7.

2. Alagona G, Bella R, Ferri R, Carnemolla A, Pappalardo A, Costanzo E, Pennisi G (2001) Transcranial magnetic stimulation in Alzheimer disease: motor cortex excitability and cognitive severity. Neurosci Lett 314:57-60.

3. Ferreri F, Pauri F, Pasqualetti P, Fini R, Dal Forno G, Rossini PM (2003) Motor cortex excitability in Alzheimer's disease: a transcranial magnetic stimulation study. Ann Neurol 53:102-8.

4. Valls-Solé J, Pascual-Leone A, Wassermann EM, Hallett M (1992) Human motor evoked responses to paired transcranial magnetic stimuli. Electroencephalogr Clin Neurophysiol 85:355-64.

5. Kujirai T, Caramia MD, Rothwell JC, Day BL, Thompson PD, Ferbert A, et al (1993) Corticocortical inhibition in human motor cortex. J Physiol 471:501-19.

6. Abbruzzese G, Trompetto C (2002) Clinical and research methods for evaluating cortical excitability. J Clin Neurophysiol 19:307-21.

7. Liepert J, Bär KJ, Meske U, Weiller C (2001) Motor cortex disinhibition in Alzheimer's disease. Clin Neurophysiol 112:1436-41.

8. Pierantozzi M, Panella M, Palmieria MG, Kocha G, Giordano A, Marciania MG, et al (2004) Different TMS patterns of intracortical inhibition in early onset Alzheimer dementia and frontotemporal dementia. Clinical Neurophysiol 115:2410–8.

11 9. Pepin JL, Bogacz D, de Pasqua V, Delwaide PJ (1999) Motor cortex inhibition is not impaired in patients with Alzheimer’s disease: evidence from paired transcranial magnetic stimulation. J Neurol Sci 170:119-23.

910. Di Lazzaro V, Oliviero A, Pilato F, Saturno E, Dileone M, Marra C, et al (2004) Motor cortex hiperexcitability to transcranial magnetic stimulation in Alzheimer’s disease. J Neurol Neurosurg Psychiatry 75:555-9.

10. Pierantozzi M, Panella M, Palmieria MG, Kocha G, Giordano A, Marciania MG, et al (2004) Different TMS patterns of intracortical inhibition in early onset Alzheimer dementia and frontotemporal dementia. Clinical Neurophysiol 115:2410–8.

11. Petersen RC, Doody R, Kurz A, Mohs RC, Morris JC, Rabins PV, et al (2001) Current concepts in mild cognitive impairment. Arch Neurol 58:1985-92.

12. Dubois B, Feldman HH, Jacova C, Dekosky ST, Barberger-Gateau P, Cummings J, et al (2007) Research criteria for the diagnosis of Alzheimer's disease: revising the NINCDS-ADRDA criteria. Lancel Neurol 6:734-46

13. Raina P, Santaguida P, Ismaila A, Patterson C, Cowan D, Levine M, et al (2008) Effectiveness of cholinesterase inhibitors and memantine for treating dementia: evidence review for a clinical practice guideline. Ann Intern Med 148:379-97.

14. Raschetti R, Albanese E, Vanacore N, Maggini M (2007) Cholinesterase inhibitors in mild cognitive impairment: a systematic review of randomised trials. PLoS Med 4:e338.

15. Zaborszky L, Duque A (2000) Local synaptic connections of basal forebrain neurons. Behav Brain Res 115:143-58.

16. Hughes CP, Berg L, Danziger WL, Coben LA and Martin RL (1982) A new clinical scale for the staging of dementia. Br J Psychiatry 140:566-72.

12 17. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (4th edn, text revision) DSM-IV-TR (2000) Washington DC: American Psychiatric Association.

18. Folstein MF, Folstein SE, McHugh PR (1975) Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12:189-98.

19. Rosen WG, Mohs RC, Davis KL (1984) A new rating scale for Alzheimer’s disease. Am J Psychiatry 141:1356-64.

20. McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM (1984) Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 34:939-44.

21. Rothman KJ. No adjustments are needed for multiple comparisons (1990) Epidemiology 1:43-6.

22. Lang N, Hasan A, Sueske E, Paulus W, Nitsche MA (2008) Cortical hypoexcitability in chronic smokers? A transcranial magnetic stimulation study. Neuropsychopharmacology 33:2517-23.

23. Di Lazzaro V, Oliviero A, Tonali PA, Marra C, Daniele A, Profice P et al (2002) Noninvasive in vivo assessment of cholinergic cortical circuits in AD using transcranial magnetic stimulation. Neurology 59:392-7.

24. Nardone R, Bergmann J, Kronbichler M, Kunz A, Klein S, Caleri F, et al (2008) Abnormal short latency afferent inhibition in early Alzheimer's disease: a transcranial magnetic demonstration. J Neural Transm 115:1557-62.

25. Sakuma K, Murakami T, Nakashima K (2007) Short latency afferent inhibition is not impaired in mild cognitive impairment. Clin Neurophysiol 118:1460-3.

13 26. Scarmeas N, Stern Y (2003) Cognitive reserve and lifestyle. J Clin Exp Neuropsychol 25:625-33.

27. Pettersson AF, Olsson E, Wahlund LO (2005) Motor function in subjects with mild cognitive impairment and early Alzheimer's disease. Dement Geriatr Cogn Disord 19:299-304.

FIGURE LEGENDS

Figure 1. pTMS recovery curve in very mild AD, at baseline, and after 3.9 (SD 0.7) months on donepezile, and in the control group. From 1 to 16 ms ISIs, a sub-threshold conditioning stimulus was utilizsed; from 20 to 300 ms ISIs, a supra-threshold conditioning stimulus was utilizsed.; dData represent the percentage of unconditioned responses. AD, : Alzheimer’s disease; Amp.:, amplitude; ISI,: interstimulus interval; pTMS,: paired-pulse transcranial magnetic stimulation;. * p < 0.05 (within AD group).

Figure 2. Correlations between test MEP recovery at 300 ms ISI and cognitive performance in 10 patients with very mild AD patients (MEP or ADAS-cog data were misseding infor two patients). MEP values (amplitude or area) represent the percentage of the conditioning response. A Hhigher score in the MMSE and lower score in the ADAS-cog indicate better cognitive performance. AD,: Alzheimer’s disease; ADAS-cog:, cognitive subscale of the Alzheimer’s Disease Assessment Scale [19]; ISI:, interstimulus interval (ms); MEP,: motor evoked potential; MMSE,: Mini-mental State TestExamination [18]; pTMS,: paired-pulse transcranial magnetic stimulation.