This single-center, randomized, single-blinded (investigator blinded, participant unblinded), controlled preliminary intervention study was conducted with a crossover design (two treatments, three periods) [3]. OH was confirmed by a 10-min tilt-table test and implied initial (blood pressure decline of > 40 mmHg systolic/20 mmHg diastolic < 15 s after standing), classic (decline of > 20 mmHg systolic/10 mmHg diastolic within 3 min of standing), and delayed OH (decline beyond 3 min) [4]. Non-neurogenic causes of OH were excluded by history, physical examination, electrocardiogram, and laboratory tests. Inclusion criteria were a diagnosis of PD, the presence of OH, sufficient mobility to perform strength training, and a minimum score of 24 out of 30 on the Montreal Cognitive Assessment (MoCA) [5].
Of 331 PD patients screened between April 2014 and October 2018 for symptoms of OH, 29 patients (16 females, mean age 69 ± 9.6 years, disease duration of 4.3 ± 3.8 years) met all inclusion criteria. Randomization was performed by random permutations in two blocks, with 15 patients assigned to group A and 14 to group B. Five patients in group A and four patients in group B dropped out after the first or second visit. Three patients had type 2 diabetes (one dropout, one each in group A and B), four patients had clinical signs of polyneuropathy (two dropouts, one each in group A and B) as an additional risk factor for autonomic dysfunction and 17 patients suffered from arterial hypertension (seven in group A and ten in group B).
Strength training was applied to both groups in a crossover design. After the baseline visit, group A had a training phase of 8 weeks, while group B served as the control group. This was followed by visit 2, an 8-week washout period, and visit 3. Group B then received strength training, while group A served as the control group. This was followed by visit 4, the washout phase, and visit 5. Group B received another 8 weeks of strength training, while group A remained the control group. The last visit took place after a total of 40 weeks.
At each of the six clinical visits between the therapy/washout phases, patients received advice on nonpharmacological treatment of OH [1, 2] and medication dosages were adjusted to minimize their influence on OH. Pressor medications (midodrine or fludrocortisone) were used and adapted in a few patients during clinical routine treatment. Tilt-table examination and heart rate variability testing (respiratory sinus arrhythmia RSA) were performed between 8 a.m. and 10 a.m., on an empty stomach without ingestion of morning medication, nicotine, or caffeine, and without wearing compression garments. The tilt-table test was performed after 10 min in the supine position, for 10 min in an upright position of 70°, with continuous noninvasive blood pressure measurement (Fan 4.1.0, Bio Sign GmbH, Ottenhofen, Germany). “Maximal systolic blood pressure change” (difference of mean systolic blood pressure during 5 min supine and minimum systolic blood pressure upright during 10 min of standing) was chosen as primary outcome variable to quantify the absolute blood pressure drop. Artifacts were removed before analysis, so that the maximum blood pressure drop did not overestimate OH. Coprimary endpoints were “maximal diastolic change”, “mean systolic/diastolic change” (difference of mean blood pressure supine and mean blood pressure during the 10-min standing period) and RSA. The standing time in which the respective minimum and maximum values were reached was expressed as the “time to reach systolic/diastolic minimum” and the “time to reach maximal heart rate (min)”. Secondary endpoints were the test results of mobility tests, cognitive tests, questionnaires, and the transcranial Doppler.
The respective therapy group received 8 weeks of training with a frequency of two sessions per week. Each 45- to 60-min training session consisted of a 20-min warm-up program (bicycle ergometer or cross-trainer) with individually set resistance, with 20 min of strength training of the leg extensor and knee flexor muscles on the leg press, leg curl, and leg extension machines with individually adjusted weights thereafter. After three sets of ten repetitions per device, the training was completed with 20 min of balance training on unstable surfaces such as balance pads. The individual strength limit of the training day of a patient was tested by gradually adjusting the weights. It was redefined for each training day. If possible, the training intensity was increased with each session, depending on the patient's performance on that day. The patient did not have to train beyond their performance limit. Patients also performed calf muscle exercises three times per week at home during the exercise periods. Exercises were not to be performed during washout periods, see supp. Figure 1 and 2.
Statistical analyses were performed using IBM SPSS Statistics 25. The significance level for all analyses was set a priori at a type I error of α = 0.05. The study was powered for at least eight patients in each group to detect an improvement of approximately 10 mmHg (range 5–15 mmHg, power, 80%) in the primary outcome. Crossover analysis of the primary and co-primary endpoints was performed using a linear mixed model (LMM), with fixed effects for treatment and period, and random patient effects. The fixed treatment effect can be interpreted as the estimated treatment difference (ES) between study groups.