BackgroundWhole-body electromyostimulation (WB-EMS) was never previously applied to Parkinson’s disease (PD) patients. This randomized controlled study aimed to find the most effective and safe WB-EMS training protocol for this population.MethodsTwenty-four subjects (age: 72.13 ± 6.20 years), were randomly assigned to three groups: a high-frequency WB-EMS strength training group (HFG) (rectangular stimulation at 85 Hz, 350 μs, 4 s stimulation/4 s rest), a low-frequency WB-EMS aerobic training group (LFG) (rectangular stimulation 7 Hz, 350 μs, with a continuous pulse duration), and an inactive control group (CG). Participants of the two experimental groups underwent 24 controlled WB-EMS training sessions, with a duration of 20 min each, during 12-week intervention. Serum growth factors (BDNF, FGF-21, NGF and proNGF), α-synuclein, physical performance and Parkinson’s Disease Fatigue Scale (PFS-16) responses were analyzed to evaluate the pre-post variation and differences among groups.ResultsSignificant interactions of Time*Groups were detected for BDNF (Time*Groups p = 0.024; Time*CG, b = −628, IC95% = −1,082/−174, p = 0.008), FGF-21 (Time*Groups p = 0.009; Time*LFG b = 1,346, IC95% = 423/2268, p = 0.005), and α-synuclein (Time*Groups p = 0.019; Time*LFG b = −1,572, IC95% = −2,952/−192, p = 0.026). Post hoc analyses and comparisons of ΔS (post–pre), performed independently for each group, showed that LFG increased serum BDNF levels (+ 203 pg/ml) and decreased α-synuclein levels (−1,703 pg/ml), while HFG showed the opposite effects (BDNF: −500 pg/ml; α-synuclein: + 1,413 pg/ml). CG showed a significant BDNF reduction over time. Both LFG and HFG showed significant improvements in several physical performance outcomes and the LFG showed better results than HFG. Concerning PFS-16, significant differences over time (b = −0.4, IC95% = −0.8/−0.0, p = 0.046) and among groups (among all groups p < 0.001) were found, and the LFG exhibited better results than the HFG (b = −1.0, IC95% = −1.3/−0.7, p < 0.001), and CG (b = −1.7, IC95% = −2.0/−1.4, p < 0.001) with this last one that worsened over time.ConclusionLFG training was the best choice for improving or maintaining physical performance, fatigue perception and variation in serum biomarkers.Clinical trial registrationhttps://www.clinicaltrials.gov/ct2/show/NCT04878679, identifier NCT04878679.