Article Text


Research paper
A multicentre prospective study of Guillain-Barré Syndrome in Japan: a focus on the incidence of subtypes
  1. Yoshiyuki Mitsui1,
  2. Susumu Kusunoki1,
  3. Kimiyoshi Arimura2,
  4. Ryuji Kaji3,
  5. Takashi Kanda4,
  6. Satoshi Kuwabara5,
  7. Masahiro Sonoo6,
  8. Kazuo Takada1
  9. and the Japanese GBS Study Group
  1. 1Faculty of Medicine, Department of Neurology, Kinki University, Osaka, Japan
  2. 2Department of Neurology, Ookatsu Hospital, Kagoshima, Japan
  3. 3Department of Neurology, Tokushima University Graduate School of Medicine, Tokushima, Japan
  4. 4Department of Neurology, Yamaguchi University Graduate School of Medicine, Yamaguchi, Japan
  5. 5Department of Neurology, Chiba University Graduate School of Medicine, Chiba, Japan
  6. 6Department of Neurology, Teikyo University School of Medicine, Tokyo, Japan
  1. Correspondence to Professor Susumu Kusunoki, 377-2 Ohno-Higashi, Osaka-Sayama, Osaka, Japan 589-8511; kusunoki-tky{at}


Objective Guillain–Barré Syndrome (GBS) is classified into the two major subtypes; acute inflammatory demyelinating polyneuropathy (AIDP) and acute motor axonal neuropathy (AMAN). Previous studies have suggested that AIDP is predominant and AMAN is rare in Western countries, whereas AMAN is not always uncommon in East Asia. We aimed to clarify the incidence of the subtypes of GBS in Japan.

Methods We performed a prospective multicentre survey over 3 years (2007–2010). Clinical and electrophysiological findings were collected from 184 patients with GBS in 23 tertiary neurology institutes. Anti-ganglioside antibodies were measured by ELISA. We also surveyed the incidence of Fisher syndrome (FS).

Results By electrodiagnostic criteria of Ho et al, patients were classified as having AIDP (40%), or AMAN (22%), or unclassified (38%). Anti-GM1 IgG antibodies were found for 47% of AMAN patients, and 18% of AIDP patients (p<0.001). There were no specific regional trends of the electrodiagnosis and anti-GM1 positivity. During the same study period, 79 patients with FS were identified; the percentage of FS cases out of all cases (FS/(GBS+FS)) was 26%.

Conclusions The frequency of GBS patients with the electrodiagnosis of AMAN by single nerve conduction studies is approximately 20% in Japan, and the AMAN pattern is closely associated with anti-GM1 antibodies. The incidence of FS appears to be much higher in Japan than in Western countries.

View Full Text

Statistics from


The concept of Guillain–Barré syndrome (GBS) changed in the 1990s due to the recognition of acute motor axonal neuropathy (AMAN) as an axonal subtype of GBS.1–3 Thus, GBS is now divided into two major subtypes: AMAN and acute inflammatory demyelinating polyradiculoneuropathy (AIDP), mainly based on neurophysiological criteria. AMAN has been associated with antecedent Campylobacter jejuni infection and autoantibodies to gangliosides, especially to GM1 and GalNAc-GD1a.4 ,5 Previous reports have suggested that AIDP is frequent and AMAN is rare in Western countries, whereas AMAN is common in Asia and central and southern America.2 ,3 ,6–18 Some regional studies in Japan7–9 ,17 suggested that the frequency of AMAN seems to be higher than in Western countries (23–48%), but this has not been shown in a large nationwide prospective survey. In the current study, we examined the electrophysiological subtypes in a large prospective cohort of the Japanese population and compared these data with those from Western countries and China.


Survey procedures

To investigate the incidence of axonal GBS, a nationwide multicentre prospective survey of GBS was conducted by the Research Committees for Neuroimmunological Diseases sponsored by the Ministry of Health, Labor and Welfare, Japan. Data for patients with GBS that fulfilled Asbury and Cornblath19 criteria were collected from 23 university hospitals or tertiary hospitals in the Japan GBS study group (see online supplementary appendix) between August 2007 and July 2010. In this period, 222 patients with GBS were treated at these hospitals. Of these patients, 184 gave informed consent for utilisation of personal data, storage and assay of biological materials for research purposes. Data for these 184 patients (male 114, female 70, age: 45.5±18.5) were collected prospectively and analysed. This is a large multicentre study in Japan.

The number of patients with Fisher syndrome (FS) was also surveyed. FS was diagnosed based on clinical symptoms characterised by acute and self-limited ophthalmoplegia, ataxia and areflexia.20 The study design was agreed upon and approved by the ethics committee of Kinki University Faculty of Medicine.

Clinical information

The following clinical information was prospectively collected for the 184 patients using predefined format: type of antecedent events (respiratory, gastrointestinal, others and none); GBS disability score21 from 0 to 6 (at nadir and at 6 months after onset: 0, healthy; 1, minor symptoms or signs, able to run; 2, able to walk 5 m independently; 3, able to walk 5 m with a walker or support; 4, bed-bound or chair-bound; 5, requiring assisted ventilation; 6, death); sensory disturbance (yes or no); deep sensory disturbance (yes or no); cranial nerve palsy (yes or no); ophthalmoplegia (yes or no); facial nerve palsy (yes or no); and oropharyngeal palsy (yes or no).


An electrophysiological study was performed within 21 days of symptom onset. If two or more studies were performed during the period in the single patients, later results were principally adopted. In the cases that second results classified into unclassified by Ho's criteria or equivocal by Hadden's criteria, first results were adopted to avoid unclassified or equivocal cases.

Motor conduction studies were performed in the median, ulnar and tibial nerves to measure the amplitude and duration of the negative peak of the compound muscle action potential (CMAP) in distal (dCMAP) and proximal (pCMAP) stimulation, conduction velocity (CV), the distal motor latency (DML), and the minimal F-wave latency. These parameters (DML, dCMAP, pCMAP and CV) are expressed as a percentage of the upper (ULN) or lower (LLN) limit of normal for the test at each hospital. Sensory nerve conduction studies were also performed in the median, ulnar, and sural nerves, but these data were not used for classification of subtypes.

Electrophysiological criteria for classification of AIDP and AMAN

Two sets of diagnostic criteria were used to discriminate between demyelination and axonal damage: the criteria of Ho et al2 (referred to as Ho's criteria) were used for comparison with data from northern China, and those of Hadden et al3 (Hadden’s criteria) were used for comparison with data from Western countries. In Ho's criteria, patients with equivocal, inexcitable or normal conditions are categorised as ‘unclassified’; therefore, patients are classified into three categories of AMAN, AIDP and unclassified in these criteria. In Hadden's criteria, patients are classified into five categories of primary axonal, primary demyelination, equivocal, inexcitable and normal.

Antiganglioside antibodies assay

IgG antibodies against gangliosides GM1 and GQ1b were measured by ELISA performed basically as described before.22 These serological data were obtained from 174 of 184 patients.

Statistical analysis

Statistical analysis was performed using SPSS V.10 software. Differences in ratios between two groups were tested for significance by χ2 test, or by Fisher exact test when the criteria for a χ2 test were not fulfilled. Analyses of numerical variables were initially performed by Kruskal–Wallis H test. When significance was found, a Mann–Whitney U test was used to determine the significance of differences between subgroups.


Electrophysiological categories

In the electrophysiological test, 73 patients (40%) met Ho's criteria for AIDP, 40 (22%) were classified as AMAN, and 71 (38%) were unclassified. Among 40 patients classified into AMAN, five (13%) was defined as acute motor and sensory axonal neuropathy (AMSAN) using electrophysiological criteria.23 The distribution of these subtypes differed significantly from those in northern China.2 (figure 1A). In the same test, 83 patients (45%) met Hadden's criteria for demyelination; 24 (13%) were classified as axonal, none as inexcitable, 16 (9%) as normal and 61 (33%) as equivocal. The distribution of these subtypes also differed significantly from those in Western countries reported by Hadden et al3 (figure 1B).

Figure 1

Comparison of data between the present study and previous studies. (A) Japan compared to China (Ho et al1) (B) Japan compared to Western countries (Hadden et al3).

Clinical and electrophysiological profiles

The clinical features of the patients are shown in tables 1 and 2. The 1st electrophysiological studies were performed at 9.2±5.3 days from symptom onset. Clinical information on antecedent events, cranial neuropathy, sensory disturbance and GBS disability score at nadir was obtained in all 184 patients, and 118 of 184 patients were also evaluated using the GBS disability score at 6 months after onset of GBS. Antiganglioside antibodies were examined in 174 patients. No patient had vaccination as an antecedent event. Two patients died during the study period, giving a mortality rate of 1.1%, and 27 patients (14.7%) required mechanical ventilation support.

Table 1

Clinical features and Hughes functional grade at nadir and 6 month after onset of GBS. Groups are classified by Ho's criteria

Table 2

Clinical features and Hughes functional grade at nadir and 6 months after onset of GBS. Groups are classified by Hadden's criteria

There was no significant difference in the type of antecedent events or GBS disability score at any time between the AMAN and AIDP classified by Ho's criteria, however, ratio of male patients, presence of overall cranial neuropathy, bulbar palsy and sensory disturbance were significantly less frequent in AMAN than in AIDP (table 1). There were no significant differences in clinical status except ratio of male patients among the subgroups classified by Hadden's criteria (table 2). An anti-GM1 antibody-positive status was more frequent in patients with AMAN compared to those with AIDP based on Ho's criteria, and also more frequent in those in the axonal subgroup compared to the demyelination subgroup based on Hadden's criteria. An anti-GQ1b antibody-positive status showed no significant difference among subgroups based on Ho's criteria and Hadden's criteria. There were no specific regional trends of the electrodiagnosis and anti-GM1 positivity in Japan.

Occurrence of FS

A total of 79 patients with FS were identified in 23 hospitals in the Japan GBS study group between August 2007 and July 2010. In the same period, there were 222 patients with GBS; therefore, the percentage of FS cases out of all cases (FS/(GBS+FS)) was 26%.


This Japanese multicentre prospective study showed an incidence of AMAN (axonal GBS) of 13–22% in electrophysiological evaluation of patients with early stage GBS. This incidence is significantly lower than that in a Chinese study,2 but higher than that in Western countries.3 In order to ensure the precise diagnosis of demyelinating or axonal type of GBS, we selected hospitals having neurologists specialised in electrophysiology. The selected hospitals were located throughout Japan. Therefore, the ratio of subtypes of GBS, demyelinating or axonal in this study, should precisely reflect the situation in Japan.

The incidences of axonal GBS found in various countries worldwide based on electrophysiological criteria are summarised in table 3. All studies classified AMAN and AIDP based on Ho's criteria, Hadden's criteria or both. Some studies included adults and children, while others focused only on children. The reasons for the different incidences of AMAN in different countries are unclear. A regional difference is apparent, since AIDP is more common in Western countries, while AMAN is more frequent in such Asian countries as China2 and Bangladesh.14 Interestingly, a report from Israel,11 which is geographically located between Europe and Asia, showed an intermediate frequency of AMAN. On the other hand, the frequency of AMAN in India,12 one of the largest countries in Asia, is lower than that in China and Bangladesh, which seems to be inconsistent with the regional trend.

Table 3

Previous studies of the incidence of AMAN (or axonal GBS) in various countries

The frequency of preceding C jejuni infection has been reported to be higher in Asian countries than in Western countries, and this was thought to be the cause of the higher incidence of the axonal type of GBS in Asia.2 However, a collaborative study in Japan and The Netherlands24 showed that the incidence of antecedent C jejuni infection in GBS in Japan was not higher than that in The Netherlands. Paradiso et al6 pointed out that 90% of children with AMAN in Argentina are from rural areas without running water, while a recent report from China showed that the incidence of AMAN was lower than that in 1995.2 ,18 Environmental factors may differ among districts within the same country, and the status of hygiene may be related to the incidence of AMAN, since the pathogenesis is thought to be correlated with gastrointestinal infection by C jejuni. Besides antecedent C jejuni infection, genetic factors might also be related to this difference, however, there is insufficient evidence. Taken together, these results suggest that regional differences mainly account for the different incidences of AMAN among various countries, but this may not fully explain this discrepancy.

The method used to classify AIDP and AMAN may be another reason for the different incidences. Ho's and Hadden's criteria were widely used to discriminate AMAN from AIDP in these studies. Hadden's criteria classify the findings of conduction block as demyelination, but Ho's criteria do not do so. Therefore, Hadden's criteria provide a wider definition of AIDP than Ho's criteria, which may partly explain the different incidences among previous studies. Furthermore, Ho's and Hadden's criteria do not take sensory conduction study into consideration. Adding data of sensory conduction study may contribute to more adequate classification of subtypes.

Another point of interest is the difference in clinical features and outcome between AMAN and AIDP. Originally, Feasby et al1described five patients with the axonal form of GBS who showed a poor outcome. Since then, the axonal form (AMAN) has been believed to have a more severe clinical course than that of AIDP. However, Hadden et al3 found no significant differences between AMAN and AIDP in their series. In our series, there were no significant differences in clinical outcome at any time between AMAN and AIDP either on Hadden's criteria or Ho's criteria. Furthermore, there were no significant differences in clinical features of antecedent events, involvement of the cranial nerve, and sensory disturbance among subtypes classified based on Hadden's criteria; however, oropharyngeal palsy and sensory disturbance were significantly less frequent in AMAN than AIDP on Ho's criteria. Our results also showed significant male predominance in AIDP or demyelinating groups. We could not find the reason for it. As the previous studies had no such gender gap, further investigation may be necessary.

The anti-GM1 antibody-positive rate was significantly higher in patients with AMAN than in those with AIDP. This trend was similar for subgroups classified using Ho's and Hadden's criteria. A large Chinese study2 and a study in Western countries3 failed to show a correlation between anti-GM1 antibody and electrodiagnosis of AMAN or axonal GBS, but accumulated clinical and experimental evidence indicates a close relationship among axonal GBS, an anti-GM1 antibody positive status, antecedent gastrointestinal events, and C jejuni. Recently, in a study in Italy, it was also reported that an anti-GM1 antibody-positive status was also correlated with AMAN.15 Our results supported the evidence of correlation between an anti-GM1 antibody-positive status and axonal damage, however, 14% of AIDP patients also had anti-GM1 antibody, indicating that the association is not so rigid. There were no significant differences in GQ1b-positive status between AIDP (demyelinating) and AMAN (axonal) subtypes.

Sekiguchi et al17 suggested that in GBS, clinical and final electrophysiological profiles are better determined by antibodies to such gangliosides as GM1, GD1a, GalNAc-GD1a and GM1b, rather than by electrodiagnosis in the early stage, and that electrodiagnosis at 3–6 weeks after GBS onset differentiates AMAN and AIDP more correct than electrodiagnosis performed in week 1 or 2. As a part of this reason, they also pointed out some patients with AIDP on the initial electrophysiological studies turned to AMAN or unclassified because the slowing of distal latency was rapidly recovered. This issue should be investigated in future studies.

In this cohort study, we also surveyed cases of FS and found that FS accounted for 26% of all GBS+FS cases. A previous study in an Italian cohort found that FS accounted for only 3% of all GBS+FS cases,25 whereas reports from a single hospital in Taiwan found this value to be 18–19%.26 ,27 In Japan, a study of 50 consecutive cases with FS indicated a value of 34%.28 These limited results suggest that FS might have a higher incidence in East Asia than in Europe.


View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:


  • For Japanese GBS Study Group see online supplementary appendix.

  • Contributions YM: drafting/revising the manuscript, analysis or interpretation of data, acquisition of data. SK: study concept or design, study supervision, obtaining funding. KA: drafting/revising the manuscript, acquisition of data. RK: drafting/revising the manuscript, acquisition of data. TK: drafting/revising the manuscript, acquisition of data. SK: drafting/revising the manuscript, acquisition of data. MS: drafting/revising the manuscript, acquisition of data. KT: and GBS epidemiological study group of Japan: acquisition of data.

  • Funding Supported in part by a Health and Labour Sciences Research Grant on Intractable Diseases (H23-Nanchi-Ippan-017) from the Ministry of Health, Labour, and Welfare, Japan, and a Scientific Research B Grant (No. 24390225) from the Ministry of Education, Culture, Sports, Science, and Technology, Japan.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval the Ethics Committee of Kinki University Faculty of Medicine.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • Neuromuscular
    Jong Seok Bae Nobuhiro Yuki Satoshi Kuwabara Jong Kuk Kim Steve Vucic Cindy S Lin Matthew C Kiernan