Bayesian Dose-Finding for Theta Burst Stimulation Tolerability: A Randomized Study Comparing Intermittent and Continuous Protocols at Distinct Prefrontal Targets

Background: Theta burst stimulation (TBS) is an efficient form of repetitive transcranial magnetic stimulation, but tolerability may depend on target and stimulation pattern, limiting translation to accelerated protocols. Objective: To estimate tolerable intensities for intermittent TBS (iTBS) over F3, approximating left dorsolateral prefrontal cortex, and continuous TBS (cTBS) over Fp1, intended to engage more anterior ventral/frontopolar circuitry, in non-treatment-seeking adults with obesity or tobacco use disorder. Methods: In an open-label randomized crossover titration study, 64 adults completed two TBS visits 7 days apart. Each visit included 3 sessions of 600 pulses, beginning at 80% resting motor threshold (RMT) with protocol-permitted escalation or de-escalation. The primary endpoint was participant-level maximum final maintained intensity. Bayesian grouped-binomial logistic regression estimated the intensity tolerated by 70% of participants (ED70), and a prespecified rule selected the highest dose with at least 80% posterior probability of meeting 70% tolerability. Results: Observed tolerability at 80% RMT was 86.4% for iTBS and 51.6% for cTBS. Primary-model ED70 was 105.5% RMT (95% credible interval [CrI], 99.0%-113.9%) for iTBS and 69.2% RMT (95% CrI, 65.0%- 73.4%) for cTBS. The recommended intensity was 100% RMT for iTBS and 60% RMT for cTBS; no cTBS dose at or above 80% RMT met criterion. cTBS produced greater immediate symptom burden, whereas 24-hour symptoms were uncommon. Conclusion: iTBS over F3 supported a future-trial design window of 90%-100% RMT, whereas cTBS over Fp1 showed a tolerability ceiling below 80% RMT. Future cTBS protocols targeting ventral prefrontal circuitry may need to move dorsally to improve tolerability.