Patients

This study was approved by the ethics committee of Kanazawa-Nishi Hospital. The patients’ consents were obtained using a written consent form. The test was performed from 2009–2010. We tested 20 Japanese patients with probable AD diagnosed with NINCDS-ADRDA criteria3 and 20 with probable DLB diagnosed with the criteria of the third international DLB workshop2 (table 1). To detect a difference of a value of 1 at the desired significance level of 0.05 and power of 0.80, 20 patients in each group were required. There were no differences in the Mini-Mental State Examination and the frontal assessment battery scores between the DLB and the AD groups. We evaluated the ratio of heart to mediastinum uptake (H/M) of the MIBG scan using single-photon emission CT. The values of H/M ratio in all the AD patients were greater than 1.70 and those of all the DLB patients were lesser than 1.50. We excluded patients with cardiovascular disease including arrhythmia, diabetes mellitus, other degenerative diseases and peripheral neuropathies. We also excluded patients not willing to participate in the study.

Testing of autonomic functions

Examinations were performed in a quiet room in which the patients relaxed and made themselves comfortable. We also offered patients a rest after the examinations.

For the SSR measurement, surface electrodes were used: the active recording electrode was placed on the palm of the hand and the reference electrode was placed on the dorsum of the same hand. SSR was induced by median nerve electrical stimulation at an amplitude of 20 mA with a 10 s interval. The waveforms of SSR appeared 1.5–2.5 s after the stimulations. Three waves were recorded for each side (right and left). The filters used in the measurement were as follows: high cut, 1 kHz; low cut, 0.1 Hz. The measured amplitude was defined as the peak-to-peak value of the recorded waves. The mean of these six amplitudes was used for the analysis.

For the HRV measurement, the A–A intervals were measured twice for 2 min with an interval of 5 min in a sitting position after a rest of 5 min. For recording and analyses of the A–A intervals, an Artett acceleration plethysmography system (U-Medica, Osaka, Japan) was utilised for analyses of the data, as described earlier.4 From the low-frequency power (LF; 0.02–0.15 Hz) and the high-frequency power (HF; 0.15–0.50 Hz), the ratio of LF to HF power (LF/HF) was calculated using the maximal entropy method (MEM). The mean of two values from MEM was used. Cut-off value was obtained from a receiver-operating characteristic (ROC) curve using DLB as the positive level and AD as the negative. The coefficient of variation of A–A intervals (CVAA) was also analysed. The examinations were performed by a clinical technician with extensive experience of such work and the results were reviewed by a special neurologist trained for working in an electrophysiological laboratory. The technician was blind to the clinical information of the patients.

For statistical analyses, the data were first tested for a normal distribution using the Shapiro-Wilk test. In the categories with a normal distribution, data were analysed for equality of variance by F test and then Student t test or Welch's t test was utilised. In the categories with a non-normal distribution, Wilcoxon's test was utilised. To detect abnormality in the SSR or HRV examinations, Fisher's exact test was used. Both the SSR and HRV examinations were performed in a quiet room and the patients were kept awake and relaxed during the procedures. The ROC curves of the data were drawn using JMP V.10.0.1 (SAS Institute Inc, Cary, North Carolina, USA). A summary of the ROC curves in this study is available as online supplemental information.