Cognitive changes following rTMS, tACS, and their combination in bipolar II disorder: an exploratory study integrating degree centrality and individualized electric field modeling

Repetitive transcranial magnetic stimulation (rTMS), transcranial alternating current stimulation (tACS), and combined tACS-rTMS may improve cognition in bipolar disorder (BD), but how individualized stimulation dose relates to neural and cognitive changes remains unclear. This secondary neuroimaging analysis included 52 euthymic participants with bipolar II disorder from a 3-week, randomized, double-blind, sham-controlled trial. Participants received combined tACS-rTMS (n = 16), rTMS (n = 14), tACS (n = 13), and sham (n = 9) conditions, with stimulation delivered to an individualized left dorsolateral prefrontal cortex target.

We integrated electric field (EF) modeling with degree centrality (DC) analysis to explore dose–network–cognition relationships. Group × time interaction effects and exploratory associations among EF intensity, changes in DC (ΔDC), and cognitive outcomes were examined using correlation, regression, and mediation analyses. Significant group × time interactions in DC were identified in all three active-versus-sham contrasts, converging in the left rolandic operculum, with additional changes in the parahippocampal and middle occipital gyri in the rTMS contrast.

Associations between ΔDC and cognition outcomes were observed but did not survive false discovery rate (FDR) correction. Nominal associations between tACS-related EF intensity and local ΔDC were observed in the tACS and combined tACS-rTMS groups. Importantly, EF intensity was significantly associated with the DC–cognition coupling index in the left caudate nucleus in the tACS group after FDR correction.

Exploratory mediation models suggested nominal indirect associations between EF intensity and cognitive change via ΔDC, neither of which survived FDR correction. In conclusion, rTMS, tACS, and combined tACS-rTMS were associated with partially distinct patterns of DC change in euthymic bipolar II disorder. Integrating individualized EF modeling with DC analysis may help characterize dose–network–cognition relationships, but larger prospective studies are needed.

We extend our sincere gratitude to all patients who participated in this study, as well as to the clinicians and medical technicians for their dedicated efforts. Research supported by the National Natural Science Foundation of China (52407261, 82201675, 82571735) and Zhejiang Provincial Key Research and Development Program (2025C01137, 2024C03098, 2025C02109). These authors contributed equally: Shuangyu Qi, Ruoyu Tang.

Department of Psychiatry, The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China Shuangyu Qi, Ruoyu Tang, Xiaomei Zhang, Hetong Zhou, Shaohua Hu & Minmin Wang Qiushi Academy for Advanced Studies, Zhejiang University, Hangzhou, China Ruoyu Tang, Yimei Gan, Tongjie Zhuang & Shaomin Zhang Nanhu Brain-computer Interface institute, Hangzhou, China The Zhejiang Key Laboratory of Precision Psychiatry, Hangzhou, China MOE Frontier Science Center for Brain Science and Brain-Machine Integration, Zhejiang University School of Medicine, Hangzhou, China Brain Research Institute of Zhejiang University, Hangzhou, China Zhejiang Engineering Center for Mathematical Mental Health, Hangzhou, China The State Key Lab of Brain-Machine Intelligence, Zhejiang University, Hangzhou, China Correspondence to Hetong Zhou, Shaohua Hu or Minmin Wang. The authors declare no competing interests. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.