Abstract
INTRODUCTION: Work-related rumination is associated with poor mental and physical health. This study aimed to develop and evaluate a personalized Bayesian optimization (pBO) algorithm, designed to tailor transcranial alternating current stimulation (tACS) parameters for UK schoolteachers experiencing elevated work-related rumination. METHODS: The algorithm adjusted tACS parameters based on individual head circumference and rumination levels. During the development phase, 80 burn-in and 319 personalized home-based sessions were conducted with 67 schoolteachers to model parameter-outcome relationships. In a preregistered, double-blind, within-participant follow-up study, 38 schoolteachers received both personalized and sham stimulation sessions in a counterbalanced order. Measures included work-related rumination, daytime sleepiness, actigraphy-based sleep fragmentation and efficiency, and side-effect ratings. RESULTS: In the algorithm development phase, higher-amplitude stimulation was associated with reduced sleep fragmentation. In the follow-up study, both personalized and sham stimulation reduced rumination, but no significant differences were observed between conditions. Higher amplitudes were associated with greater reductions in daytime sleepiness. Post-stimulation changes in sleepiness and ruminations were associated only at higher amplitudes in the personalized condition and were not observed following sham stimulation. Side-effect severity did not differ significantly between conditions. CONCLUSIONS: Home-based neurostimulation interventions are feasible and well tolerated. While personalized stimulation did not outperform sham, the findings identify current amplitude as a key factor influencing daytime sleepiness and highlight the need to further refine personalization algorithms and stimulation strategies to optimize effectiveness.