Digital phenotyping: towards replicable findings with comprehensive assessments and integrative models in bipolar disorders.
Ulrich W Ebner-PriemerEsther MühlbauerAndreas B NeubauerHolger HillFabrice BeierPhilip S SantangeloPhilipp RitterNikolaus KleindienstMichael BauerFlorian SchmiedekEmanuel SeverusPublished in: International journal of bipolar disorders (2020)
Combining gold-standard categorical expert ratings with dimensional self and expert ratings resulted in two latent outcomes (mania and depression) with statistically meaningful factor loadings that dynamically varied over 299 days. Latent digital phenotypes of sleep and activity were associated with same-day latent manic psychopathology, suggesting that psychopathological alterations in bipolar disorders relate to domains (latent variables of sleep and activity) and not only to specific behaviors (such as the number of declined incoming calls). The identification of latent psychopathological outcomes that dimensionally vary on a daily basis will enable to empirically determine which combination of digital phenotypes at which days prior to an upcoming episode are viable as digital prodromal predictors.