Mood instability in patients with unipolar depression measured using smartphones and the association with measures of wellbeing, recovery and functioning.

OBJECTIVE: While mood instability is strongly linked to depression, its ramifications remain unexplored. In patients diagnosed with unipolar depression (UD), our objective was to investigate the association between mood instability, calculated based on daily smartphone-based patient-reported data on mood, and functioning, quality of life, perceived stress, empowerment, rumination, recovery, worrying and wellbeing. METHODS: Patients with UD completed daily smartphone-based self-assessments of mood for 6 months, making it possible to calculate mood instability using the Root Mean Squared Successive Difference (rMSSD) method. A total of 59 patients with UD were included. Data were analyzed using mixed effects regression models. RESULTS: There was a statistically significant association between increased mood instability and increased perceived stress (adjusted model: B: 0.010, 95% CI: 0.00027; 0.021, p = 0.044), and worrying (adjusted model: B: 0.0060, 95% CI: 0.000016; 0.012, p = 0.049), and decreased quality of life (adjusted model: B: -0.0056, 95% CI: -0.011; -0.00028, p = 0.039), recovery (adjusted model: B: -0.032, 95% CI: -0.0059; -0.00053, p = 0.019) and wellbeing. There were no statistically significant associations between mood instability and functioning, empowerment, and rumination (p's >0.09). CONCLUSION: These findings underscore the significant influence of mood instability on patients' daily lives. Identification of mood fluctuations offer potential insights into the trajectory of the illness in these individuals.

Using digital phenotyping to classify bipolar disorder and unipolar disorder - exploratory findings using machine learning models.

The aims were to investigate 1) differences in smartphone-based data on phone usage between bipolar disorder (BD) and unipolar disorder (UD) and 2) by using machine learning models, the sensitivity, specificity, and AUC of the combined smartphone data in classifying BD and UD. Daily smartphone-based self-assessments of mood and same-time passively collected smartphone data on smartphone usage was available for six months. A total of 64 patients with BD and 74 patients with UD were included. Patients with BD during euthymic states compared with UD in euthymic states had a lower number of incoming phone calls/ day (B: -0.70, 95%CI: -1.37; -0.70, p=0.040). Patients with BD during depressive states had a lower number of incoming and outgoing phone calls/ day as compared with patients with UD in depressive states. In classification by using machine learning models, 1) overall (regardless of the affective state), patients with BD were classified with an AUC of 0.84, which reduced to 0.48 when using a leave-one-patient-out crossvalidation (LOOCV) approach; similarly 2) during a depressive state, patients with BD were classified with an AUC of 0.86, which reduced to 0.42 with LOOCV; 3) during a euthymic state, patients with BD were classified with an AUC of 0.87, which reduced to 0.46 with LOOCV. While digital phenotyping shows promise in differentiating between patients with BD and UD, it highlights the challenge of generalizing to unseen individuals. It should serve as an complement to comprehensive clinical evaluation by clinicians.

Graph neural network interatomic potential ensembles with calibrated aleatoric and epistemic uncertainty on energy and forces.

Inexpensive machine learning (ML) potentials are increasingly being used to speed up structural optimization and molecular dynamics simulations of materials by iteratively predicting and applying interatomic forces. In these settings, it is crucial to detect when predictions are unreliable to avoid wrong or misleading results. Here, we present a complete framework for training and recalibrating graph neural network ensemble models to produce accurate predictions of energy and forces with calibrated uncertainty estimates. The proposed method considers both epistemic and aleatoric uncertainty and the total uncertainties are recalibrated post hoc using a nonlinear scaling function to achieve good calibration on previously unseen data, without loss of predictive accuracy. The method is demonstrated and evaluated on two challenging, publicly available datasets, ANI-1x (Smith et al. J. Chem. Phys., 2018, 148, 241733.) and Transition1x (Schreiner et al. Sci. Data, 2022, 9, 779.), both containing diverse conformations far from equilibrium. A detailed analysis of the predictive performance and uncertainty calibration is provided. In all experiments, the proposed method achieved low prediction error and good uncertainty calibration, with predicted uncertainty correlating with expected error, on energy and forces. To the best of our knowledge, the method presented in this paper is the first to consider a complete framework for obtaining calibrated epistemic and aleatoric uncertainty predictions on both energy and forces in ML potentials.

The effect of smartphone-based monitoring and treatment including clinical feedback versus smartphone-based monitoring without clinical feedback in bipolar disorder: the SmartBipolar trial-a study protocol for a randomized controlled parallel-group trial.

INTRODUCTION: A substantial proportion of patients with bipolar disorder experience daily subsyndromal mood swings, and the term "mood instability" reflecting the variability in mood seems associated with poor prognostic factors, including impaired functioning, and increased risk of hospitalization and relapse. During the last decade, we have developed and tested a smartphone-based system for monitoring bipolar disorder. The present SmartBipolar randomized controlled trial (RCT) aims to investigate whether (1) daily smartphone-based outpatient monitoring and treatment including clinical feedback versus (2) daily smartphone-based monitoring without clinical feedback or (3) daily smartphone-based mood monitoring only improves mood instability and other clinically relevant patient-related outcomes in patients with bipolar disorder. METHODS AND ANALYSIS: The SmartBipolar trial is a pragmatic randomized controlled parallel-group trial. Patients with bipolar disorder are invited to participate as part of their specialized outpatient treatment for patients with bipolar disorder in Mental Health Services in the Capital Region of Denmark. The included patients will be randomized to (1) daily smartphone-based monitoring and treatment including a clinical feedback loop (intervention group) or (2) daily smartphone-based monitoring without a clinical feedback loop (control group) or (3) daily smartphone-based mood monitoring only (control group). All patients receive specialized outpatient treatment for bipolar disorder in the Mental Health Services in the Capital Region of Denmark. The trial started in March 2021 and has currently included 150 patients. The outcomes are (1) mood instability (primary), (2) quality of life, self-rated depressive symptoms, self-rated manic symptoms, perceived stress, satisfaction with care, cumulated number and duration of psychiatric hospitalizations, and medication (secondary), and (3) smartphone-based measures per month of stress, anxiety, irritability, activity, and sleep as well as the percentage of days with presence of mixed mood, days with adherence to medication and adherence to smartphone-based self-monitoring. A total of 201 patients with bipolar disorder will be included in the SmartBipolar trial. ETHICS AND DISSEMINATION: The SmartBipolar trial is funded by the Capital Region of Denmark and the Independent Research Fund Denmark. Ethical approval has been obtained from the Regional Ethical Committee in The Capital Region of Denmark (H-19067248) as well as data permission (journal number: P-2019-809). The results will be published in peer-reviewed academic journals, presented at scientific meetings, and disseminated to patients' organizations and media outlets. TRIAL REGISTRATION: Trial registration number: NCT04230421. Date March 1, 2021. Version 1.

Mood instability and activity/energy instability in patients with bipolar disorder according to day-to-day smartphone-based data - An exploratory post hoc study.

BACKGROUND: Alterations and instability in mood and activity/energy has been associated with impaired functioning and risk of relapse in bipolar disorder. The present study aimed to investigate whether mood instability and activity/energy instability are associated, and whether these instability measures are associated with stress, quality of life and functioning in patients with bipolar disorder. METHODS: Data from two studies were combined for exploratory post hoc analyses. Patients with bipolar disorder provided smartphone-based evaluations of mood and activity/energy levels from day-to-day. In addition, information on functioning, perceived stress and quality of life was collected. A total of 316 patients with bipolar disorder were included. RESULTS: A total of 55,968 observations of patient-reported smartphone-based data collected from day-to-day were available. Regardless of the affective state, there was a statistically significant positive association between mood instability and activity/energy instability in all models (all p-values < 0.0001). There was a statistically significant association between mood and activity/energy instability with patient-reported stress and quality of life (e.g., mood instability and stress: B: 0.098, 95 % CI: 0.085; 0.11, p < 0.0001), and between mood instability and functioning (B: 0.045, 95 % CI: 0.0011; 0.0080, p = 0.010). LIMITATIONS: Findings should be interpreted with caution since the analyses were exploratory and post hoc by nature. CONCLUSION: Mood instability and activity/energy instability is suggested to play important roles in the symptomatology of bipolar disorder. This highlight that monitoring and identifying subsyndromal inter-episodic fluctuations in symptoms is clinically recommended. Future studies investigating the effect of treatment on these measures would be interesting.

Irritability in bipolar disorder and unipolar disorder measured daily using smartphone-based data: An exploratory post hoc study.

OBJECTIVE: To investigate (i) the proportions of time with irritability and (ii) the association between irritability and affective symptoms and functioning, stress, and quality of life in patients with bipolar disorder (BD) and unipolar depressive disorder (UD). METHODS: A total of 316 patients with BD and 58 patients with UD provided self-reported once-a-day data on irritability and other affective symptoms using smartphones for a total of 64,129 days with observations. Questionnaires on perceived stress and quality of life and clinical evaluations of functioning were collected multiple times during the study. RESULTS: During a depressive state, patients with UD spent a significantly higher proportion of time with presence of irritability (83.10%) as compared with patients with BD (70.27%) (p = 0.045). Irritability was associated with lower mood, activity level and sleep duration and with increased stress and anxiety level, in both patient groups (p-values<0.008). Increased irritability was associated with impaired functioning and increased perceived stress (p-values<0.024). In addition, in patients with UD, increased irritability was associated with decreased quality of life (p = 0.002). The results were not altered when adjusting for psychopharmacological treatments. CONCLUSIONS: Irritability is an important part of the symptomatology in affective disorders. Clinicians could have focus on symptoms of irritability in both patients with BD and UD during their course of illness. Future studies investigating treatment effects on irritability would be interesting.

Transition1x - a dataset for building generalizable reactive machine learning potentials.

Machine Learning (ML) models have, in contrast to their usefulness in molecular dynamics studies, had limited success as surrogate potentials for reaction barrier search. This is primarily because available datasets for training ML models on small molecular systems almost exclusively contain configurations at or near equilibrium. In this work, we present the dataset Transition1x containing 9.6 million Density Functional Theory (DFT) calculations of forces and energies of molecular configurations on and around reaction pathways at the ωB97x/6-31 G(d) level of theory. The data was generated by running Nudged Elastic Band (NEB) with DFT on 10k organic reactions of various types while saving intermediate calculations. We train equivariant graph message-passing neural network models on Transition1x and cross-validate on the popular ANI1x and QM9 datasets. We show that ML models cannot learn features in transition state regions solely by training on hitherto popular benchmark datasets. Transition1x is a new challenging benchmark that will provide an important step towards developing next-generation ML force fields that also work far away from equilibrium configurations and reactive systems.

Cheap Turns Superior: A Linear Regression-Based Correction Method to Reaction Energy from the DFT.

Workflows to predict chemical reaction networks based on density functional theory (DFT) are prone to systematic errors in reaction energy due to the extensive use of cheap DFT exchange-correlation functionals to limit computational cost. Recently, machine learning-based models are increasingly applied to mitigate this problem. However, machine learning models require systems similar to trained data, and the models often perform poorly for out-of-distribution systems. Here, we present a simple bond-based correction method that improves the accuracy of DFT-derived reaction energies. It is based on linear regression, and the correction terms for each bond are derived from reactions among the QM9 data set. We demonstrate the effectiveness of this method with three DFT functionals in three different rungs of Jacob's ladder. The simple correction method is effective for all rungs but especially so for the cheapest PBE functional. Finally, we applied the correction method to a few reactions with molecules significantly different from those in the QM9 data set that was used to fit the linear regression model. Once corrected by this method, we found that the DFT reaction energies for such out-of-distribution reactions are within 0.05 eV of the G4MP2 method.

Differences in mobility patterns according to machine learning models in patients with bipolar disorder and patients with unipolar disorder.

BACKGROUND: It is essential to differentiate bipolar disorder (BD) from unipolar disorder (UD) as the course of illness and treatment guidelines differ between the two disorders. Measurements of activity and mobility could assist in this discrimination. AIMS: 1) To investigate differences in smartphone-based location data between BD and UD, and 2) to investigate the sensitivity, specificity, and AUC of combined location data in classifying BD and UD. METHODS: Patients with BD and UD completed smartphone-based self-assessments of mood for six months, along with same-time passively collected smartphone data on location reflecting mobility patterns, routine and location entropy (chaos). A total of 65 patients with BD and 75 patients with UD were included. RESULTS: A total of 2594 (patients with BD) and 2088 (patients with UD) observations of smartphone-based location data were available. During a depressive state, compared with patients with UD, patients with BD had statistically significantly lower mobility (e.g., total duration of moves per day (eB 0.74, 95% CI 0.57; 0.97, p = 0.027)). In classification models during a depressive state, patients with BD versus patients with UD, there was a sensitivity of 0.70 (SD 0.07), a specificity of 0.77 (SD 0.07), and an AUC of 0.79 (SD 0.03). LIMITATIONS: The relative low symptom severity in the present study may have contributed to the magnitude of the AUC. CONCLUSION: Mobility patterns derived from mobile location data is a promising digital diagnostic marker in discriminating between patients with BD and UD.

Discriminating between patients with unipolar disorder, bipolar disorder, and healthy control individuals based on voice features collected from naturalistic smartphone calls.

BACKGROUND: It is of crucial importance to be able to discriminate unipolar disorder (UD) from bipolar disorder (BD), as treatments, as well as course of illness, differ between the two disorders. AIMS: To investigate whether voice features from naturalistic phone calls could discriminate between (1) UD, BD, and healthy control individuals (HC); (2) different states within UD. METHODS: Voice features were collected daily during naturalistic phone calls for up to 972 days. A total of 48 patients with UD, 121 patients with BD, and 38 HC were included. A total of 115,483 voice data entries were collected (UD [n = 16,454], BD [n = 78,733], and HC [n = 20,296]). Patients evaluated symptoms daily using a smartphone-based system, making it possible to define illness states within UD and BD. Data were analyzed using random forest algorithms. RESULTS: Compared with BD, UD was classified with a specificity of 0.84 (SD: 0.07)/AUC of 0.58 (SD: 0.07) and compared with HC with a sensitivity of 0.74 (SD: 0.10)/AUC = 0.74 (SD: 0.06). Compared with BD during euthymia, UD during euthymia was classified with a specificity of 0.79 (SD: 0.05)/AUC = 0.43 (SD: 0.16). Compared with BD during depression, UD during depression was classified with a specificity of 0.81 (SD: 0.09)/AUC = 0.48 (SD: 0.12). Within UD, compared with euthymia, depression was classified with a specificity of 0.70 (SD 0.31)/AUC = 0.65 (SD: 0.11). In all models, the user-dependent models outperformed the user-independent models. CONCLUSIONS: The results from the present study are promising, but as reflected by the low AUCs, does not support that voice features collected during naturalistic phone calls at the current state of art can be implemented in clinical practice as a supplementary and assisting tool. Further studies are needed.

Voice analyses using smartphone-based data in patients with bipolar disorder, unaffected relatives and healthy control individuals, and during different affective states.

BACKGROUND: Voice features have been suggested as objective markers of bipolar disorder (BD). AIMS: To investigate whether voice features from naturalistic phone calls could discriminate between (1) BD, unaffected first-degree relatives (UR) and healthy control individuals (HC); (2) affective states within BD. METHODS: Voice features were collected daily during naturalistic phone calls for up to 972 days. A total of 121 patients with BD, 21 UR and 38 HC were included. A total of 107.033 voice data entries were collected [BD (n = 78.733), UR (n = 8004), and HC (n = 20.296)]. Daily, patients evaluated symptoms using a smartphone-based system. Affective states were defined according to these evaluations. Data were analyzed using random forest machine learning algorithms. RESULTS: Compared to HC, BD was classified with a sensitivity of 0.79 (SD 0.11)/AUC = 0.76 (SD 0.11) and UR with a sensitivity of 0.53 (SD 0.21)/AUC of 0.72 (SD 0.12). Within BD, compared to euthymia, mania was classified with a specificity of 0.75 (SD 0.16)/AUC = 0.66 (SD 0.11). Compared to euthymia, depression was classified with a specificity of 0.70 (SD 0.16)/AUC = 0.66 (SD 0.12). In all models the user dependent models outperformed the user independent models. Models combining increased mood, increased activity and insomnia compared to periods without performed best with a specificity of 0.78 (SD 0.16)/AUC = 0.67 (SD 0.11). CONCLUSIONS: Voice features from naturalistic phone calls may represent a supplementary objective marker discriminating BD from HC and a state marker within BD.

The effect of smartphone-based monitoring and treatment on the rate and duration of psychiatric readmission in patients with unipolar depressive disorder: The RADMIS randomized controlled trial.

BACKGROUND: Patients with unipolar depressive disorder are frequently hospitalized, and the period following discharge is a high-risk-period. Smartphone-based treatments are receiving increasing attention among researchers, clinicians, and patients. We aimed to investigate whether a smartphone-based monitoring and treatment system reduces the rate and duration of readmissions, more than standard treatment, in patients with unipolar depressive disorder following hospitalization. METHODS: We conducted a pragmatic, investigator-blinded, randomized controlled trial. The intervention group received a smartphone-based monitoring and treatment system in addition to standard treatment. The system allowed patients to self-monitor symptoms and access psycho-educative information and cognitive modules. The patients were allocated a study-nurse who, based on the monitoring data, guided and supported them. The control group received standard treatment. The trial lasted six months, with outcome assessments at 0, 3, and 6 months. RESULTS: We included 120 patients with unipolar depressive disorder (ICD-10). Intention-to-treat analyses showed no statistically significant differences in time to readmission (Log-Rank p=0.9) or duration of readmissions (B=-16.41,95%CI:-47.32;25.5,p=0.3) (Primary outcomes). There were no differences in clinically rated depressive symptoms (p=0.6) or functioning (p=0.1) (secondary outcomes). The intervention group had higher levels of recovery (B=7,29, 95%CI:0.82;13,75,p=0.028) and a tendency towards higher quality of life (p=0.07), wellbeing (p=0,09) satisfaction with treatment (p=0.05) and behavioral activation (p=0.08) compared with the control group (tertiary outcomes). LIMITATIONS: Patients and study-nurses were unblinded to allocation. CONCLUSIONS: We found no effect of the intervention on primary or secondary outcomes. In tertiary outcomes, patients in the intervention group reported higher levels of recovery compared to the control group.

Reducing the rate of psychiatric re-admissions in bipolar disorder using smartphones-The RADMIS trial.

OBJECTIVES: The MONARCA I and II trials were negative but suggested that smartphone-based monitoring may increase quality of life and reduce perceived stress in bipolar disorder (BD). The present trial was the first to investigate the effect of smartphone-based monitoring on the rate and duration of readmissions in BD. METHODS: This was a randomized controlled single-blind parallel-group trial. Patients with BD (ICD-10) discharged from hospitalization in the Mental Health Services, Capital Region of Denmark were randomized 1:1 to daily smartphone-based monitoring including a feedback loop (+ standard treatment) or to standard treatment for 6 months. Primary outcomes: the rate and duration of psychiatric readmissions. RESULTS: We included 98 patients with BD. In ITT analyses, there was no statistically significant difference in rates (hazard rate: 1.05, 95% CI: 0.54; 1.91, p = 0.88) or duration of readmission between the two groups (B: 3.67, 95% CI: -4.77; 12.11, p = 0.39). There was no difference in scores on the Hamilton Depression Rating Scale (B = -0.11, 95% CI: -2.50; 2.29, p = 0.93). The intervention group had higher scores on the Young Mania Rating Scale (B: 1.89, 95% CI: 0.0078; 3.78, p = 0.050). The intervention group reported lower levels of perceived stress (B: -7.18, 95% CI: -13.50; -0.86, p = 0.026) and lower levels of rumination (B: -6.09, 95% CI: -11.19; -1.00, p = 0.019). CONCLUSIONS: Smartphone-based monitoring did not reduce rate and duration of readmissions. There was no difference in levels of depressive symptoms. The intervention group had higher levels of manic symptoms, but lower perceived stress and rumination compared with the control group.

Daily mobility patterns in patients with bipolar disorder and healthy individuals.

BACKGROUND: Alterations in energy and activity in bipolar disorder (BD) differ between affective states and compared with healthy control individuals (HC). Measurements of activity could discriminate between BD and HC and in the monitoring of affective states within BD. The aims were to investigate differences in 1) passively collected smartphone-based location data (location data) between BD and HC, and 2) location data in BD between affective states. METHODS: Daily, patients with BD and HC completed smartphone-based self-assessments of mood for up to nine months. Location data reflecting mobility patterns, routine and location entropy was collected daily. A total of 46 patients with BD and 31 HC providing daily data was included. RESULTS: A total of 4,859 observations of smartphone-based self-assessments of mood and mobility patterns were available from patients with BD and 1,747 observations from HC. Patients with BD had lower location entropy compared with HC (B= -0.14, 95% CI= -0.24; -0.034, p=0.009). Patients with BD during a depressive state were less mobile compared with a euthymic state. Patients with BD during an affective state had lower location entropy compared with a euthymic state (p<0.0001). The AUC of combined location data was rather high in classifying patients with BD compared with HC (AUC: 0.83). LIMITATIONS: Individuals willing to use smartphones for daily self-monitoring may represent a more motivated group. CONCLUSION: Alterations in location data reflecting mobility patterns may be a promising measure of illness and illness activity in patients with BD and may be used to monitor the effects of treatments.

Smartphone-based activity measurements in patients with newly diagnosed bipolar disorder, unaffected relatives and control individuals.

BACKGROUND: In DSM-5 activity is a core criterion for diagnosing hypomania and mania. However, there are no guidelines for quantifying changes in activity. The objectives of the study were (1) to investigate daily smartphone-based self-reported and automatically-generated activity, respectively, against validated measurements of activity; (2) to validate daily smartphone-based self-reported activity and automatically-generated activity against each other; (3) to investigate differences in daily self-reported and automatically-generated smartphone-based activity between patients with bipolar disorder (BD), unaffected relatives (UR) and healthy control individuals (HC). METHODS: A total of 203 patients with BD, 54 UR, and 109 HC were included. On a smartphone-based app, the participants daily reported their activity level on a scale from -3 to + 3. Additionally, participants owning an android smartphone provided automatically-generated data, including step counts, screen on/off logs, and call- and text-logs. Smartphone-based activity was validated against an activity questionnaire the International Physical Activity Questionnaire (IPAQ) and activity items on observer-based rating scales of depression using the Hamilton Depression Rating scale (HAMD), mania using Young Mania Rating scale (YMRS) and functioning using the Functional Assessment Short Test (FAST). In these analyses, we calculated averages of smartphone-based activity measurements reported in the period corresponding to the days assessed by the questionnaires and rating scales. RESULTS: (1) Smartphone-based self-reported activity was a valid measure according to scores on the IPAQ and activity items on the HAMD and YMRS, and was associated with FAST scores, whereas the majority of automatically-generated smartphone-based activity measurements were not. (2) Daily smartphone-based self-reported and automatically-generated activity correlated with each other with nearly all measurements. (3) Patients with BD had decreased daily self-reported activity compared with HC. Patients with BD had decreased physical (number of steps) and social activity (more missed calls) but a longer call duration compared with HC. UR also had decreased physical activity compared with HC but did not differ on daily self-reported activity or social activity. CONCLUSION: Daily self-reported activity measured via smartphone represents overall activity and correlates with measurements of automatically generated smartphone-based activity. Detecting activity levels using smartphones may be clinically helpful in diagnosis and illness monitoring in patients with bipolar disorder. Trial registration clinicaltrials.gov NCT02888262.

Daily self-reported and automatically generated smartphone-based sleep measurements in patients with newly diagnosed bipolar disorder, unaffected first-degree relatives and healthy control individuals.

OBJECTIVES: (1) To investigate daily smartphone-based self-reported and automatically generated sleep measurements, respectively, against validated rating scales; (2) to investigate if daily smartphone-based self-reported sleep measurements reflected automatically generated sleep measurements and (3) to investigate the differences in smartphone-based sleep measurements between patients with bipolar disorder (BD), unaffected first-degree relatives (UR) and healthy control individuals (HC). METHODS: We included 203 patients with BD, 54 UR and 109 HC in this study. To investigate whether smartphone-based sleep calculated from self-reported bedtime, wake-up time and screen on/off time reflected validated rating scales, we used the Pittsburgh Sleep Quality Index (PSQI) and sleep items on the Hamilton Depression Rating Scale 17-item (HAMD-17) and the Young Mania Rating Scale (YMRS). FINDINGS: (1) Self-reported smartphone-based sleep was associated with the PSQI and sleep items of the HAMD and the YMRS. (2) Automatically generated smartphone-based sleep measurements were associated with daily self-reports of hours slept between 12:00 midnight and 06:00. (3) According to smartphone-based sleep, patients with BD slept less between 12:00 midnight and 06:00, with more interruption and daily variability compared with HC. However, differences in automatically generated smartphone-based sleep were not statistically significant. CONCLUSION: Smartphone-based data may represent measurements of sleep patterns that discriminate between patients with BD and HC and potentially between UR and HC. CLINICAL IMPLICATION: Detecting sleep disturbances and daily variability in sleep duration using smartphones may be helpful for both patients and clinicians for monitoring illness activity. TRIAL REGISTRATION NUMBER: clinicaltrials.gov (NCT02888262).

Daily estimates of clinical severity of symptoms in bipolar disorder from smartphone-based self-assessments.

Currently, the golden standard for assessing the severity of depressive and manic symptoms in patients with bipolar disorder (BD) is clinical evaluations using validated rating scales such as the Hamilton Depression Rating Scale 17-items (HDRS) and the Young Mania Rating Scale (YMRS). Frequent automatic estimation of symptom severity could potentially help support monitoring of illness activity and allow for early treatment intervention between outpatient visits. The present study aimed (1) to assess the feasibility of producing daily estimates of clinical rating scores based on smartphone-based self-assessments of symptoms collected from a group of patients with BD; (2) to demonstrate how these estimates can be utilized to compute individual daily risk of relapse scores. Based on a total of 280 clinical ratings collected from 84 patients with BD along with daily smartphone-based self-assessments, we applied a hierarchical Bayesian modelling approach capable of providing individual estimates while learning characteristics of the patient population. The proposed method was compared to common baseline methods. The model concerning depression severity achieved a mean predicted R2 of 0.57 (SD = 0.10) and RMSE of 3.85 (SD = 0.47) on the HDRS, while the model concerning mania severity achieved a mean predicted R2 of 0.16 (SD = 0.25) and RMSE of 3.68 (SD = 0.54) on the YMRS. In both cases, smartphone-based self-reported mood was the most important predictor variable. The present study shows that daily smartphone-based self-assessments can be utilized to automatically estimate clinical ratings of severity of depression and mania in patients with BD and assist in identifying individuals with high risk of relapse.

Forecasting Mood in Bipolar Disorder From Smartphone Self-assessments: Hierarchical Bayesian Approach.

BACKGROUND: Bipolar disorder is a prevalent mental health condition that is imposing significant burden on society. Accurate forecasting of symptom scores can be used to improve disease monitoring, enable early intervention, and eventually help prevent costly hospitalizations. Although several studies have examined the use of smartphone data to detect mood, only few studies deal with forecasting mood for one or more days. OBJECTIVE: This study aimed to examine the feasibility of forecasting daily subjective mood scores based on daily self-assessments collected from patients with bipolar disorder via a smartphone-based system in a randomized clinical trial. METHODS: We applied hierarchical Bayesian regression models, a multi-task learning method, to account for individual differences and forecast mood for up to seven days based on 15,975 smartphone self-assessments from 84 patients with bipolar disorder participating in a randomized clinical trial. We reported the results of two time-series cross-validation 1-day forecast experiments corresponding to two different real-world scenarios and compared the outcomes with commonly used baseline methods. We then applied the best model to evaluate a 7-day forecast. RESULTS: The best performing model used a history of 4 days of self-assessment to predict future mood scores with historical mood being the most important predictor variable. The proposed hierarchical Bayesian regression model outperformed pooled and separate models in a 1-day forecast time-series cross-validation experiment and achieved the predicted metrics, R2=0.51 and root mean squared error of 0.32, for mood scores on a scale of -3 to 3. When increasing the forecast horizon, forecast errors also increased and the forecast regressed toward the mean of data distribution. CONCLUSIONS: Our proposed method can forecast mood for several days with low error compared with common baseline methods. The applicability of a mood forecast in the clinical treatment of bipolar disorder has also been discussed.

Is smartphone-based mood instability associated with stress, quality of life, and functioning in bipolar disorder?

OBJECTIVE: Mood instability in patients with bipolar disorder has been associated with impaired functioning and risk of relapse. The present study aimed to investigate whether increased mood instability is associated with increased perceived stress and impaired quality of life and functioning in patients with bipolar disorder. METHODS: A total of 84 patients with bipolar disorder used a smartphone-based self-monitoring system on a daily basis for 9 months. Data on perceived stress, quality of life, and clinically rated functioning were collected at five fixed time points for each patient during follow-up. A group of 37 healthy individuals served as a control comparison of perceived stress, quality of life, and psychosocial functioning. RESULTS: The majority of patients presented in full or partial remission. As hypothesized, mood instability was significantly associated with increased perceived stress (B: 10.52, 95% CI: 5.25; 15.77, P < 0.0001) and decreased quality of life (B: -12.17, 95% CI. -19.54; -4.79, P < 0.0001) and functioning (B: -12.04, 95% CI: -19.08; -4.99, P < 0.0001) in patients with bipolar disorder. There were no differences in mood instability according to prescribed psychopharmacological treatment. Compared with healthy individuals, patients reported substantially increased perceived stress and experienced decreased quality of life and decreased functioning based on researcher-blinded evaluation. CONCLUSION: Mood instability in bipolar disorder is associated with increased perceived stress and decreased quality of life and functioning even during full or partial remission. There is a need to monitor and identify subsyndromal inter-episodic symptoms. Future studies investigating the effect of treatment on mood instability are highly warranted.

Differences in mood instability in patients with bipolar disorder type I and II: a smartphone-based study.

BACKGROUND: Mood instability in bipolar disorder is associated with a risk of relapse. This study investigated differences in mood instability between patients with bipolar disorder type I and type II, which previously has been sparingly investigated. METHODS: Patients with bipolar disorder type I (n = 53) and type II (n = 31) used a daily smartphone-based self-monitoring system for 9 months. Data in the present reflect 15.975 observations of daily collected smartphone-based data on patient-evaluated mood. RESULTS: In models adjusted for age, gender, illness duration and psychopharmacological treatment, patients with bipolar disorder type II experienced more mood instability during depression compared with patients with bipolar disorder type I (B: 0.27, 95% CI 0.007; 0.53, p = 0.044), but lower intensity of manic symptoms. Patients with bipolar disorder type II did not experience lower mean mood or higher intensity of depressive symptoms compared with patients with bipolar disorder type I. CONCLUSIONS: Compared to bipolar disorder type I, patients with bipolar disorder type II had higher mood instability for depression. Clinically it is of importance to identify these inter-episodic symptoms. Future studies investigating the effect of treatment on mood instability measures are warranted. Trial registration NCT02221336.