Deriving symptom networks from digital phenotyping data in serious mental illness

Abstract

Background Symptoms of serious mental illness are multidimensional and often interact in complex ways. Generative models offer value in elucidating the underlying relationships that characterise these networks of symptoms. Aims In this paper we use generative models to find unique interactions of schizophrenia symptoms as experienced on a moment-by-moment basis. Method Self-reported mood, anxiety and psychosis symptoms, self-reported measurements of sleep quality and social function, cognitive assessment, and smartphone touch screen data from two assessments modelled after the Trail Making A and B tests were collected with a digital phenotyping app for 47 patients in active treatment for schizophrenia over a 90-day period. Patients were retrospectively divided up into various non-exclusive subgroups based on measurements of depression, anxiety, sleep duration, cognition and psychosis symptoms taken in the clinic. Associated transition probabilities for the patient cohort and for the clinical subgroups were calculated using state transitions between adjacent 3-day timesteps of pairwise survey domains. Results The three highest probabilities for associated transitions across all patients were anxiety-inducing mood (0.357, P < 0.001), psychosis-inducing mood (0.276, P < 0.001), and anxiety-inducing poor sleep (0.268, P < 0.001). These transition probabilities were compared against a validation set of 17 patients from a pilot study, and no significant differences were found. Unique symptom networks were found for clinical subgroups. Conclusions Using a generative model using digital phenotyping data, we show that certain symptoms of schizophrenia may play a role in elevating other schizophrenia symptoms in future timesteps. Symptom networks show that it is feasible to create clinically interpretable models that reflect the unique symptom interactions of psychosis-spectrum illness. These results offer a framework for researchers capturing temporal dynamics, for clinicians seeking to move towards preventative care, and for patients to better understand their lived experience.

Digital tools for the assessment of pharmacological treatment for depressive disorder: State of the art

“…These markers can highlight a stressful moment for an individual at risk or provide an indication of resilience and can inform pharmacological choices. Therefore, handling and understanding the patterns of parameters and investigating interactions between pre-existing traits and acute changes, are important to allow accurate profiling and identification of individuals at risk, relevant depression subtypes, or endophenotypes and inform treatment choices ( Hays et al, 2020 ;Onnela, 2021 ;Ploubidis et al, 2020 ). In addition to the identification of individuals at risk, the individualized opportunities to train skills improving resilience to cope with difficult times ( Southwick and Charney, 2012 ), as well as treating these prodromal behavioural changes up front, e.g., insomnia ( Cheng et al, 2019 ), are also important assets of digital phenotyping when it comes to preventing depression ( Ebert and Cuijpers, 2018 ).…”

Remote cognitive assessment in severe mental illness: a scoping review

“…Notable comprehensive remote batteries that reported acceptable psychometric properties included the Brief Assessment of Cognition 39 , My Cognition Quotient 35 , Online Neurocognitive Assessments 36 , and Screen for Cognitive Assessment in Psychiatry 48 . Some individual tasks also showed valid, sensitive, and/or reliable remote administration, particularly the Jewel Trail Making Task from the mindLAMP smartphone application, used in three studies 49 – 51 .…”

Reinventing mental health care in youth through mobile approaches: Current status and future steps

“…In some studies, digital phenotyping has been compared with gold-standard research methods to diagnose mental health disorders ( Torous et al, 2019 ; Hays et al, 2020 ; Carpenter et al, 2021 ; Melcher et al, 2021 ; Moshe et al, 2021 ). However, more research should be carried out to highlight what EMAs and digital phenotyping can add to the information collected with gold-standard measures, in terms of explained variance in outcome variables and predicting power.…”