Neural microstates in real-world behaviour captured on the smartphone

Microstates are periods of quasi-stability in large-scale neural networks, and they are ubiquitous spanning sleep, rest and behaviour. One possibility is that the temporal and spatial features of these states vary from person to person contributing to inter-individual behavioural differences. Another possibility is that the microstates support momentary behavioural needs and as a result their properties may vary with behavioural fluctuations in the same person. Here, we leverage the history of smartphone touchscreen interaction dynamics and combined smartphone and EEG recordings (N = 61), to address if and how microstates contribute to real-world behaviour. We find that microstates measured when at rest correlated with the history ([~]30 days) of smartphone interaction dynamics captured in the real world, indicating their role in the behaviour. During smartphone behaviour, the configuration of microstates varied over time. However, this variance was largely unrelated to the rich behavioural dynamics, as in similar temporal and spatial features were observed during rapid and slow smartphone interactions. We show that in the real world microstates do not respond to the fluctuations of the ongoing behaviour. Still, as microstates measured at rest are correlated to the inter-individual behavioural differences captured on the smartphone, we propose that microstates exert a top-down influence to broadly orchestrate real-world behaviour.