Brain activity during dual-task standing in older adults

A secondary analysis was completed on data from a multi-site study testing the effects of single exposures to transcranial direct current stimulation targeting different brain regions on dual-task standing performance in older adults [18]. That study was approved by the Institutional Review Boards of Hebrew SeniorLife and Tel Aviv Sourasky Medical Center and conducted in accordance with the principles of the Declaration of Helsinki. All participants provided written informed consent at the beginning of their first study visit. For the current analysis, we focused on the pre-stimulation dual-task standing assessments that were completed during four laboratory visits each separated by approximately one week. The analysis was further limited to data from the participants tested at the Hebrew SeniorLife site because only this site included EEG recordings.

Subjects

Thirty older adults were recruited from the community and tested at Hebrew SeniorLife’s Hinda and Arthur Marcus Institute for Aging Research (Boston, MA, U.S). Participants were included in the study if they were persons aged 65 and older, able to read, write and communicate in English, and able to stand and walk without use of an assistive device. Exclusion criteria included a Montreal Cognitive Assessment (MoCA) [19] score < 24, self-reported presence of neurodegenerative conditions such as Parkinson’s disease or multiple sclerosis, self-report of acute illness, injury or other unstable medical condition and hospitalization within the past three months; self-reported active cancer or other terminal diseases; any report of severe lower-extremity arthritis, pain or orthopedic problems that would likely affect gait or standing balance; physician-diagnosis of peripheral neuropathy, or other peripheral neuromuscular disease; use of antipsychotics, anti-seizure, or other neuroactive medications; any report or physician-diagnosis of schizophrenia or other psychiatric illness.

Experimental protocol

All participants completed an initial screening and baseline visit. After providing informed consent, the research staff recorded the individual’s demographics, height, weight, medical history, and medications. The MoCA was then completed. Interested and eligible participants were enrolled in the study and completed the Timed Up and Go test of mobility. They were then scheduled to complete four additional in-person experimental visits during which EEG and postural sway were recorded during a dual-task standing assessment.

At the beginning of each experimental visit, participants were outfitted with wearable motion sensors (APDM, Portland, OR) [26] and a 32-channel EEG system (Starstim Enobio 32, Neuroelectrics Inc., Cambridge, MA). The motion sensors were secured to the left instep, the right instep, and the lumbar spine using Velcro straps. The EEG system was secured in place using a fitted Neoprene cap aligned with prefabricated holes corresponding to 10–20 EEG system. The electrodes were prepared with conductive electrode gel (SuperVisc 100 gr. HighViscosity Electrolyte-Gel for active electrodes, Brain Products). Data acquisition was communicated wirelessly through the EEG system connected to a laptop computer and recorded through Neuroscan software designed by Neuroelectrics, Inc.

The dual-task standing assessment was comprised of two 60-s trials of standing in each of two conditions: eyes open (i.e., single-task) and eyes open while performing a serial subtraction task (i.e., dual-task) [20]. The serial subtraction task involved audibly counting backward by 3’s from a random 3-digit number between 200 and 999 provided immediately before the trial. During each trial, participants were instructed to keep arms at their side and feet shoulder-width apart. Foot placement was traced in the first trial and this tracing was used in all subsequent trials. Before each trial, participants were reminded to avoid extraneous movements and focus their vision on a small “X” drawn on a wall at eye-level approximately three meters away.

Postural sway analysis

The instrumented SWAY test was used to assess postural sway with the APDM Mobility Lab software (APDM, Portland, OR). A laptop wirelessly collected data from the sensors at a sampling frequency of 128 Hz and was processed by algorithms developed by the manufacturer to quantify postural sway parameters. Measures of postural sway included mean total sway area (m2/s4), sway velocity (m/s), and sway path (m/s2) were derived from each trial. The dual-task cost to each outcome was computed by calculating the absolute change in each outcome from the single to the dual-task condition [21, 22].

EEG processing

EEG data was collected with a sampling rate of 500 Hz (Hz) and data was preprocessed and analyzed using the software, CARTOOL [23]. Raw EEG data files were converted by MATLAB into readable format for the CARTOOL software. The data were filtered with DC/Baseline Removal, a Butterworth High Pass Filter at 1 Hz, a Butterworth Low Pass Filter at 80 Hz, a Notch filter at 60 Hz, and then exported into a binary file format to be processed in MATLAB for independent component analysis (ICA) to remove eye blinks and movement artifacts [24]. Electrode impedances were kept below 5 kΩ in all recordings and electrode sites. All electrodes were referred to linked ear lobes, and a ground electrode was attached to the center of the forehead. Noisy channels were identified by visual inspection (standard deviation qualitatively higher than the other measured channels) and interpolated using the nearest-neighbor spline method. Trials with more than eight artifact channels were rejected (8% of trials were rejected from quiet standing and 14% of trials were rejected from dual-tasking). On average, standing conditions had 3.8 ± 0.6 channels rejected whereas dual-tasking conditions had 5.6 ± 2.2 channels rejected. Data were epoched in consecutive two-second windows and any window with noisy data was rejected from the final analysis. Finally, all remaining windows were concatenated into a continuous time series that were used for frequency analysis. Subjects with fewer than 8 s of data were not included in the final frequency analysis [25].

EEG frequency analysis was completed using the CARTOOL spectral analysis function. Spectral analysis is the change of signal power, recorded in microvolts, in the frequency domain. Frequency records the number of oscillations per second in specific bands of interest [26]. We calculated the mean absolute power density (μV2/Hz) of alpha (8–16 Hz), theta (4–7 Hz), and beta (18–32 Hz) by using fast Fourier transformation (FFT) by ‘Neuromapping-3,55’ (MBN, Russia). EEG frequencies alpha, theta, beta, power, and theta/beta power ratio were examined with region-of-interest (ROI) analyses. In accordance with Bohle et al. [27], we demarcated six ROIs, anterior left (AL) (F7, Fp1, F3, FC1, FC5, and AF3), central left (CL) (C3, CP1, CP5, and T7), posterior left (PL) (P7, P3, PO1, and PO3), anterior right (AR) (F4, Fp2, F8, FC2, FC6, and AF4), central right (CR) (C4, T8, CP2, and CP6), and posterior right (PR) (P8, P4, O2, and PO4).

Statistical analysis

Statistical analyses were performed using JMP Pro 14 software (SAS Institute, Cary, NC). Descriptive statistics (i.e., mean and standard deviation (SD)) were used to summarize the demographic characteristics of participants and study outcomes. Shapiro–Wilk tests and histograms were used to examine if the EEG outcomes and postural sway metrics were normally distributed. Variables that were not normally distributed were log transformed prior to modeling. For the EEG outcomes, we excluded outliers defined by data points more than four standard deviations away from the mean of that variable.

First, as participants completed the same protocol during four different visits, we examined the test–retest reliability of EEG-derived metrics during single and dual-task conditions using intraclass correlation coefficient (ICC) analysis [28]. ICC was interpreted as follows: greater than 0.70 was excellent, 0.60 to 0.69 was good, 0.40 to 0.59 was fair, and less than 0.40 was poor [29].

Second, the effect of dual-tasking on EEG power across frequency bands was examined using mixed model repeated measures analysis. Data points from all four visits were included in the analysis. Mixed effects models included condition (single vs dual tasking) was included as a fixed effect and a random intercept for subject. Separate analyses were completed for each dependent variable; that is, absolute alpha, theta, and beta power, and theta/beta ratio.

The relationship between each EEG outcome from each region of interest and each postural sway metric during standing and dual-tasking were assessed by using Pearson’s r correlations for the means of the 4 visits within each participant. The results were interpreted as follows: greater than 0.70 was strong, 0.50 to 0.70 was moderate, and 0.30 to 0.50 was weak [30]. The significance level was set to p < 0.05 for all analyses.

留言 (0)

沒有登入
gif