Predicting Mood Disturbance Severity with Mobile Phone Keystroke Metadata: A BiAffect Digital Phenotyping Study.

BACKGROUND: Mood disorders are common and associated with significant morbidity and mortality. Better tools are needed for their diagnosis and treatment. Deeper phenotypic understanding of these disorders is integral to the development of such tools. This study is the first effort to use passively collected mobile phone keyboard activity to build deep digital phenotypes of depression and mania. OBJECTIVE: The objective of our study was to investigate the relationship between mobile phone keyboard activity and mood disturbance in subjects with bipolar disorders and to demonstrate the feasibility of using passively collected mobile phone keyboard metadata features to predict manic and depressive signs and symptoms as measured via clinician-administered rating scales. METHODS: Using a within-subject design of 8 weeks, subjects were provided a mobile phone loaded with a customized keyboard that passively collected keystroke metadata. Subjects were administered the Hamilton Depression Rating Scale (HDRS) and Young Mania Rating Scale (YMRS) weekly. Linear mixed-effects models were created to predict HDRS and YMRS scores. The total number of keystrokes was 626,641, with a weekly average of 9791 (7861), and that of accelerometer readings was 6,660,890, with a weekly average 104,076 (68,912). RESULTS: A statistically significant mixed-effects regression model for the prediction of HDRS-17 item scores was created: conditional R2=.63, P=.01. A mixed-effects regression model for YMRS scores showed the variance accounted for by random effect was zero, and so an ordinary least squares linear regression model was created: R2=.34, P=.001. Multiple significant variables were demonstrated for each measure. CONCLUSIONS: Mood states in bipolar disorder appear to correlate with specific changes in mobile phone usage. The creation of these models provides evidence for the feasibility of using passively collected keyboard metadata to detect and monitor mood disturbances.

Equivalent linear change in cognition between individuals with bipolar disorder and healthy controls over 5 years.

OBJECTIVES: Cognitive dysfunction is a key feature of bipolar disorder (BD). However, not much is known about its temporal stability, as some studies have demonstrated a neurodegenerative model in BD while others have shown no change in cognitive functioning over time. Building upon our prior work, which examined the natural course of executive functioning, the current study aimed to investigate the natural course of memory, emotion processing, and fine motor dexterity over a 5-year period in BD and healthy control (HC) samples. METHODS: Using a 5-year longitudinal cohort, 90 individuals with BD and 17 HCs were administered a battery of neuropsychological tests at study baseline and at 1 and 5 years after study entry that captured four areas of cognitive performance: visual memory, auditory memory, emotion processing, and fine motor dexterity. RESULTS: Latent growth curve modeling showed no group differences in the slopes of any of the cognitive factors between the BD and HC groups. Age at baseline was negatively associated with visual memory, emotion processing, and fine motor dexterity. Education level was positively associated with auditory and visual memory and fine motor. Female gender was negatively associated with emotion processing. CONCLUSIONS: Extending our prior work on longitudinal evaluation of executive functioning, individuals with BD show similar linear change in other areas of cognitive functioning including memory, emotion processing, and fine motor dexterity as compared to unaffected HCs. Age, education, and gender may have some differential effects on cognitive changes.

A Smartphone App to Monitor Mood Symptoms in Bipolar Disorder: Development and Usability Study.

BACKGROUND: There is considerable scientific interest in finding new and innovative ways to capture rapid fluctuations in functioning within individuals with bipolar disorder (BD), a severe, recurrent mental disorder associated with frequent shifts in symptoms and functioning. The use of smartphones can provide valid and real-world tools for use in measurement-based care and could be used to inform more personalized treatment options for this group, which can improve standard of care. OBJECTIVE: We examined the feasibility and usability of a smartphone to capture daily fluctuations in mood within BD and to relate daily self-rated mood to smartphone use behaviors indicative of psychomotor activity or symptoms of the illness. METHODS: Participants were 26 individuals with BD and 12 healthy control individuals who were recruited from the Prechter Longitudinal Study of BD. All were given a smartphone with a custom-built app and prompted twice a day to complete questions of mood for 28 days. The app automatically and unobtrusively collected phone usage data. A poststudy satisfaction survey was also completed. RESULTS: Our sample showed a very high adherence rate to the daily momentary assessments (91% of the 58 prompts completed). Multivariate mixed effect models showed that an increase in rapid thoughts over time was associated with a decrease in outgoing text messages (ß=-.02; P=.04), and an increase in impulsivity self-ratings was related to a decrease in total call duration (ß=-.29; P=.02). Participants generally reported positive experiences using the smartphone and completing daily prompts. CONCLUSIONS: Use of mobile technology shows promise as a way to collect important clinical information that can be used to inform treatment decision making and monitor outcomes in a manner that is not overly burdensome to the patient or providers, highlighting its potential use in measurement-based care.

Decreased working memory capacity among individuals with a mood disorder who have increased metabolic burden.

BACKGROUND: Individuals with mood disorders experience a higher rate of obesity than the general population, putting them at risk for poorer outcomes. The relationship between obesity and a core feature of the mood disorders, neurocognition, is less understood. We examined the interaction of obesity as indexed by body mass index (BMI) and working memory performance in a large sample of individuals with bipolar disorder (BD), major depressive disorder (MDD), and healthy controls (HC). METHODS: Participants with BD (n = 133), MDD (n = 78), and HC (n = 113) (age range 18-40) completed a spatial working memory (SWM) task that included three-graded increases in the number of target locations. Participants were subdivided by BMI classification into six diagnostic-BMI (BMI groups: Normal Weight, Overweight/Obese) subgroups. Performance on the task was indexed by number of errors within each difficulty level. RESULTS: The number of errors, across all groups, increased with task difficulty. There was an interaction between errors and diagnostic-BMI group. Post-hoc analyses indicated that while the Normal Weight-BD group did not differ in performance from the other groups, the Overweight/Obese-BD group performed significantly worse than HC groups. LIMITATIONS: Metabolic effects of psychotropic medications due to the naturalistic nature of the study, younger age of the MDD sample, and utilizing self-reported indicators of obesity may limit generalizability. CONCLUSIONS: Individuals with BD with increased metabolic burden exhibit increased working memory errors than non-psychiatric controls who also have increased metabolic burden. Future work could address prevention and amelioration of such difficulties to reduce associated functional morbidity.