Psychomotor Slowing in Psychosis and Inhibitory Repetitive Transcranial Magnetic Stimulation

Key Points Question Can inhibitory transcranial magnetic stimulation ameliorate psychomotor slowing in psychosis? Findings In this 4-arm randomized clinical trial including 88 patients, 15 sessions of 1-Hz repetitive transcranial magnetic stimulation (rTMS) on the supplementary motor area led to response in significantly more patients than intermittent theta burst stimulation (iTBS), sham, or no treatment. Most of the patients in the waiting group responded to delayed-onset 1-Hz rTMS. Meaning The findings indicate that add-on inhibitory rTMS may be an effective treatment for patients with psychosis and psychomotor slowing; further studies are needed to assess neural changes associated with this treatment.

eMethods 1. Sample size estimation eMethods 2. Trial registration eTable 1. Inclusion/Exclusion criteria eMethods 3. Coils eMethods 4. Secondary outcome rating scale subscores eMethods 5. Actigraphy eMethods 6. Coin rotation task eMethods 7. Randomization and allocation concealment eResults eTable 2. Dropout reasons eTable 3.Primary and secondary outcomes according to treatment arm for ITT with LOCF and covariates eTable 4. Primary and secondary outcomes according to treatment arms with LOCF (raw) eTable 5. Post-hoc-p-values of Time with SIDAK correction eTable 6. Post-hoc-p-values the interaction between Time & Treatment arm eTable 7. Follow-ups with LOCF and covariates eTable 8.Primary and secondary outcomes according to treatment arm of completers with covariates eTable 9.Primary and secondary outcomes according to treatment arm with completers (raw) eTable 10.Follow ups of completers with covariates eTable 11.Side-effects in the waiting group eFigure 1. Course of anhedonia eFigure 2. Course of global functioning eTable 12. Baseline variables in the three sample versions eTable 13. Results of the primary outcomes in n=88 (ITT) vs. n=97 (with baseline), covaried for sex, baseline and treatment medication eReferences This supplemental material has been provided by the authors to give readers additional information about their work.

eMethods eMethods 1. Sample size estimation
The main effects of the 3-week rTMS interventions on psychomotor slowing (SRRS scores) was calculated in a repeated measures design including 2 assessment time points (Baseline, week 3) and four groups (lf-rTMS, iTBS, sham, waiting group-lf-rTMS).Assuming a moderate effect size (f = 0.23) as indicated by our pilot data in a repeated measures ANOVA with moderate correlation between time points (0.5), a power of 0.95 and an alpha = 0.05, we needed 88 patients (22 per group).

eMethods 2. Trial registration
The trial was registered 25 days after enrolling the first participant.On March 12,   2019, IRB communicated final approval of the protocol.On March 22, 2019, the first participant provided informed consent.First rTMS treatment started on March 27, 2019.Trial registration was conducted on April 16, 2019.At this time, the first patient had just completed week 3 measurements, another two patients were receiving rTMS, while another two patients had been enrolled but withdrawn consent before starting the intervention.Thus, when the trial was registered, a total of 3 (3.4%) of the ITT population had been enrolled in the study.
Trial registration at clinicaltrials.govincluded a minimal revision of the measurement organization that was conducted on April 25, 2019.No changes were made to the study design or outcomes after that.
Trial registration allowed entering the planned two primary outcomes (proportion of responders at week 3 and change in SRRS total scores from baseline to week 3).
The IRB protocol only allowed for one primary outcome, i.e., change in SRRS total scores from baseline to week 3. Therefore, in the IRB protocol, the proportion of responders was stated as first of the secondary outcomes.This discrepancy between IRB and clinical trial registration was present from April 2019 onwards.In the presentation of the manuscript, we adhere to the clinical trial registration.For the real TMS-stimulations we used the MCF-B70 coil with transducer head dimensions of 180x116x50/67, magnetic field of 12kT/s in coil center at 20mm distance from coil surface and penetration depth (70V/m) of 34.0mm.In contrast, for sham stimulations, we used the MCF-P-B65 coil with transducer head dimensions of 172x94x53.These coils look almost identical and have identical sound emissions.
However, the sham coil has no magnetic emissions.

eMethods 4. Secondary outcome rating scale subscores
Next to the total scores of the Positive And Negative Syndrome Scale (PANSS) 1 , the Brief Negative Symptom Scale (BNSS) 2 , the Bush-Francis Catatonia Rating Scale (BFCRS) 3 , we calculated secondary outcomes for subscores of those scales.

eMethods 5. Actigraphy
The change in total physical activity during waking hours was measured with wrist actigraphy on the non-dominant arm (Move4 by movisens GmbH, Karlsruhe, Germany) 5 .Participants wore the instrument for > 24 consecutive hours.To this end, we used the activity level (activity counts/h of wake time) 6 , which has been calculated in multiple of our previous studies 5,[7][8][9][10] .Wake time of the day was confirmed based on inspection of the raw actigraphy data and crosschecking with the sleep/activity protocol participants provided.

eMethods 6. Coin rotation task
Manual dexterity of both hands was tested using the coin rotation task 11 .We have applied this task before in samples of patients with schizophrenia and healthy subjects [12][13][14] Participants were asked to rotate a Swiss 50-Rappen coin similar to the size of a US dime, between the first three fingers at a maximum speed with each hand during 3 trails of 10s, while being videotaped.Videos were analyzed by raters blind to allocation and time point of assessment.The first trial was a practice trial whereas the remaining two trials' scores were averaged and considered for the analysis.The following formula was used to calculate the score of each trial: CR score=half turns−[(coin drops×.10)×halfturns] 11,14

eMethods 7. Randomization and allocation concealment
Prior to the first participant enrolled, we used a web-based tool "research randomizer" to perform permutated block randomization.Lists with random order of the four arms were generated.This was done to ensure that all four treatment arms receive the same number of participants.Permutated block randomization is preferred in sample sizes < 100 as in our trial 15 .Only the principal investigator (SW) had access to the randomization protocol that was stored on a login-secured computer.
Furthermore, group allocation of the participants was performed according to the block randomization lists by the principal investigator immediately after participants signed the consent form -and before baseline assessments.This way, information from baseline assessments were not available during the group allocation process.
The principal investigator was not involved in participant screening or recruitment.
The principal investigator informed the part of the team who conducted the rTMS treatments (only one person for each participant) about group allocation.The psychiatrists who performed the assessments (DA, DBG, and AK) had no access to the randomization lists or treatment allocation process.
Allocation concealment was assured by keeping the treatment allocation written in sealed envelopes, locked at the lab office.In case of emergency, e.g.serious adverse event, the principal investigator could have broken the code and informed the clinical team about the treatment allocation.However, the study would have stopped in this participant.Fortunately, SAEs have never occurred during the study.
See also IRB protocol chapter 6 for information on randomization, blinding, and unblinding.
In treatment arms 1 Hz, iTBS, and sham participants, raters, and the mental health professionals delivering treatment were blind to the stimulation used.Conversely, the participants randomized to the patient group knew that they were not receiving rTMS during the first three weeks.They were told to receive real rTMS in weeks 4-6.In the participant information it was stated that after the waiting period, patients would receive rTMS.But as the clinical trial registration is publicly available, participants may have noted that the waiting group would receive inhibitory stimulation through week 4-6. .

eTable 1 .
Inclusion/Exclusion criteria Detailed inclusion and exclusion criteria.Inclusion criteria -18-60 years old -Schizophrenia spectrum disorders according to DSM-5 with psychomotor slowing (SRRS score ≥ 15) Exclusion criteria -Substance abuse or dependence other than nicotine -Past or current medical or neurological conditions associated with impaired or aberrant movement, such as brain tumors, stroke, M. Parkinson, M. Huntington, dystonia, or severe head trauma with subsequent loss of consciousness -Epilepsy or other convulsions -History of any hearing problems or ringing in the ears -Standard exclusion criteria for MRI scanning and TMS; e.g.metal implants, claustrophobia -Any TMS treatment in the past 3 months -Pregnancy or breastfeeding eMethods 3. Coils All stimulations were delivered using either MagPro X100 including MagOption or MagPro R30 with theta burst option, both manufactured by Tonica Electronik A/S, Denmark and distributed by Magventure A/S, Denmark.Each machine was consistently used at the same building of the department (MagPro X100 at Bolligenstrasse 111, MagPro R30 at Murtenstrasse 21).Patients were treated at the rTMS machine located closest to their inpatient stay.

Primary and secondary outcomes until week 3 LOCF eTable 3. Primary and secondary outcomes according to treatment arm for ITT with LOCF and covariates. Estimated Marginal Mean ± SE
SNS © 2024 Walther S et al.JAMA Psychiatry.© 2024 Walther S et al.JAMA Psychiatry.eTable 4. Primary and secondary outcomes according to treatment arms with LOCF (raw).©2024 Walther S et al.JAMA Psychiatry.© 2024 Walther S et al.JAMA Psychiatry.Post-

Primary and secondary outcomes according to treatment arm with completers (raw).
© 2024 Walther S et al.JAMA Psychiatry.

Side-effects in the waiting group
Note: * There were too few patients to provide an exact number without compromising data identifiability requirements.

Baseline variables in the three sample versions.
SRRS bl: baseline value of the Salpetriere Retardation Rating Scale score; blOLZ: baseline olanzapine equivalents in mg; blDE: baseline diazepam equivalents in mg; mOLZ: mean olanzapine equivalents in mg; mDE: mean diazepam equivalents in mg.

Results of the primary outcomes in n=88 (ITT) vs. n=97 (with baseline), covaried for sex, baseline and treatment medication.
Note: ITT n = 88 includes patients receiving ≥ 1 stimulations (ITT).n = 97 includes all patients with baseline assessments.All missing data were computed with last observation carried forward (LOCF).