Dopaminergic manipulation modulates frequency-specific EEG connectivity patterns: Evidence from a single dose drug challenge study

Alina J. Preuss^1,2Renate de Bock^3,4Amatya J. Mackintosh⁵Simon Schuster⁶Philipp Sterzer^3,4Jonas Obleser^2,7Stefan Borgwardt^1,2Christina Andreou^1,3*

• ¹Translational Psychiatry, Department of Psychiatry and Psychotherapy, University of Lübeck, Lübeck, Germany
• ²Center of Brain, Behavior, and Metabolism (CBBM), University of Lübeck, Lübeck, Germany
• ³Universitat Basel, Basel, Switzerland
• ⁴Universitare Psychiatrische Kliniken Basel, Basel, Switzerland
• ⁵University of Basel, Faculty of Psychology, Department Clinical Psychology and Epidemiology, Basel, Switzerland
• ⁶Asklepios Klinik Barmbek, Hamburg, Germany
• ⁷Department of Psychology, University of Lübeck, Lübeck, Germany

Dysregulation in resting-state connectivity in electroencephalogram (EEG) and dysregulation in the dopamine system have both been repeatedly observed in patients with psychotic disorders. It is unclear whether dopaminergic drugs, used to treat psychotic disorders, induce treatment-relevant connectivity changes in the brain. Investigating the effects of dopaminergic drugs in healthy participants could provide crucial insights into how dopaminergic modulation directly influences resting-state connectivity. In a randomized controlled crossover study including 58 healthy participants, we examined the effects of L-dopa (dopamine pre-cursor), haloperidol (dopamine D2 antagonist), and placebo on frequency-specific connectivity patterns (intrinsic coupling modes; ICMs). Source-space connectivity between brain regions was calculated for phase ICMs via multivariate interaction measure, and for envelope ICMs via orthogonalized power correlation. Phase ICMs showed a significant linear increase (from L-dopa to placebo to haloperidol) in the delta frequency, and a significant linear decrease in the low-beta frequency. Gamma phase ICMs were significantly reduced in both drug conditions compared to placebo. Orthogonalized power correlation showed a significant linear decrease (from L-dopa to placebo to haloperidol), in the alpha and low-beta frequency. Our findings in the gamma band support the long-standing hypothesis that dopaminergic action on large-scale neural networks follows an inverted U-shaped curve. The linear pattern of our findings in the delta, alpha, and low-beta frequency bands reflect possible relevant effects for psychotic symptoms. However, our findings are difficult to relate to existing findings in patients. Our results highlight the need for further prospective studies in patients to better understand the relationship between dopaminergic effects and ICMs.