Personalized transcranial alternating current stimulation (tACS) and physical therapy to treat motor and cognitive symptoms in Parkinson's disease: A randomized cross-over trial

Abstract

Abnormal cortical oscillations are markers of Parkinson's Disease (PD). Transcranial alternating current stimulation (tACS) can modulate brain oscillations and possibly impact on behaviour. Mapping of cortical activity (prevalent oscillatory frequency and topographic scalp distribution) may provide a personalized neurotherapeutic target and guide non-invasive brain stimulation. This is a cross-over, double blinded, randomized trial. Electroencephalogram (EEG) from participants with PD referred to Specialist Clinic, University Hospital, were recorded. TACS frequency and electrode position were individually defined based on statistical comparison of EEG power spectra maps with normative data from our laboratory. Stimulation frequency was set according to the EEG band displaying higher power spectra (with beta excess on EEG map, tACS was set at 4 Hz; with theta excess, tACS was set at 30 Hz). Participants were randomized to tACS or random noise stimulation (RNS), 5 days/week for 2-weeks followed by ad hoc physical therapy. EEG, motor (Unified Parkinson's Disease Rating Scale-motor: UPDRS III), neuropsychological (frontal, executive and memory tests) performance and mood were measured before (T₀), after (T₁) and 4-weeks after treatment (T₂). A linear model with random effects and Wilcoxon test were used to detect differences. Main results include a reduction of beta rhythm in theta-tACS vs. RNS group at T₁ over right sensorimotor area (p = .014) and left parietal area (p = .010) and at T₂ over right sensorimotor area (p = .004) and left frontal area (p = .039). Bradykinesia items improved at T₁ (p = .002) and T₂ (p = .047) compared to T₀ in the tACS group. In the tACS group the Montréal Cognitive Assessment (MoCA) improved at T₂ compared with T₁ (p = .049). Individualized tACS in PD improves motor and cognitive performance. These changes are associated with a reduction of excessive fast EEG oscillations.

Keywords: Electroencephalography; Neurophysiology; Neurotherapeutic target; Non-invasive brain stimulation (NIBS); Unified Parkinson's disease rating scale (UPDRS III).

Fig. 1

Schematic representation of workflow of EEG analysis at single subject level. After pre-processing, FFT was applied to non-overlapping 2-s epochs and then averaged across epochs. Power spectral density was estimated for all frequencies and the relative power (%) was computed for delta (1–4 Hz), theta (4.5–7.5 Hz), alpha1 (8–10 Hz), alpha2 (10.5–12.5 Hz) and beta (13–30 Hz) frequency ranges, producing different topographical maps. Relative powers of a single subject were statistically compared (z-test) to relative powers of a control group. From this comparison, we derived the frequency band and the site of stimulation, based on statistically significant different electrodes.

Fig. 2

Experimental design.

Fig. 3

Example of statistical comparison. Statistical maps derived from two subjects with PD vs. control group: one subject stimulated in theta range (Pt 4) and one in beta range (Pt 8). (Left) Participant 4 shows higher beta activity, compared to controls, over FC1 and C3. He was stimulated in theta range over C3 (black circle at T₀). Beta activity reduction was observed after real stimulation but not after sham stimulation. (Right) Participant 8 shows higher theta activity, compared to controls, mainly over left frontal areas. He was stimulated in beta range over F3 (black circle at T₀). No significant modifications were observed after both real and sham stimulation.

Fig. 4

Study flow diagram.