This confirmatory study aimed to examine whether we can foresee recurrence of depressive symptoms using personalized modeling of rises in restlessness.

Participants were formerly depressed patients (N = 41) in remission who (gradually) discontinued antidepressants. Participants completed five smartphone-based Ecological Momentary Assessments (EMA) a day, for a period of 4 months, yielding a total of 21 180 observations. Statistical Process Control by means of Exponentially Weighted Moving Average (EWMA) control charts was used to detect rises in the EMA item ‘I feel restless’, for each individual separately.

An increase in restlessness was detected in 68.3% of the participants with recurring depressive symptoms, and in 26.3% of those who stayed in remission (Fisher's exact test p = 0.01, sensitivity was 68.3%, specificity was 73.7%). In the participants with a recurrence and an increase in restlessness, this increase could be detected in the prodromal phase of depression in 93.3% of the cases and at least a month before the onset of the core symptoms of depression in 66.7% of the cases.

Restlessness is a common prodromal symptom of depression. The sensitivity and specificity of the EWMA charts was at least as good as prognostic models based on cross-sectional patient characteristics. An advantage of the current idiographic method is that the EWMA charts provide real-time personalized insight in a within-person increase in early signs of depression, which is key to alert the right patient at the right time.

Predicting future states of psychopathology such as depressive episodes has been a hallmark initiative in mental health research. Dynamical systems theory has proposed that rises in certain ‘early warning signals’ (EWSs) in time-series data (e.g. auto-correlation, temporal variance, network connectivity) may precede impending changes in disorder severity. The current study investigates whether rises in these EWSs over time are associated with future changes in disorder severity among a group of patients with major depressive disorder (MDD).

Thirty-one patients with MDD completed the study, which consisted of daily smartphone-delivered surveys over 8 weeks. Daily positive and negative affect were collected for the time-series analyses. A rolling window approach was used to determine whether rises in auto-correlation of total affect, temporal standard deviation of total affect, and overall network connectivity in individual affect items were predictive of increases in depression symptoms.

Results suggested that rises in auto-correlation were significantly associated with worsening in depression symptoms (r = 0.41, p = 0.02). Results indicated that neither rises in temporal standard deviation (r = −0.23, p = 0.23) nor in network connectivity (r = −0.12, p = 0.59) were associated with changes in depression symptoms.

This study more rigorously examines whether rises in EWSs were associated with future depression symptoms in a larger group of patients with MDD. Results indicated that rises in auto-correlation were the only EWS that was associated with worsening future changes in depression.

For patients with bipolar disorder, early recognition of impending mood episodes is crucial to enable timely intervention. Longitudinal digital mood monitoring using ecological momentary assessment (EMA) enable prospective study of early warning signals (EWS) in momentary affective estates prior to symptom transitions.

The present study examined in a unique longitudinal EMA data set whether EWS prospectively signal transitions to manic or depressive episodes.

Twenty bipolar type I/II patients completed EMA questionnaires five times a day for four months (average 491 observations per person), as well as weekly symptom questionnaires concerning depressive (Quick Inventory for Depressive Symptomatology) and manic (Altman Self-Rating Mania Scale) symptoms. Weekly data was used to determine transitions (i.e., abrupt increase in symptoms). Prior to these transitions, EWS (autocorrelation at lag-1 and standard deviation) were calculated in moving windows over 17 affective EMA states. Kendall’s tau was calculated to detect significant rises in the EWS indicator prior to the transition.

Eleven patients reported one or two transitions to a mood episode. All transitions were preceded by at least one EWS. Average sensitivity for detecting EWS was slightly higher for manic episodes (36%) than for depressive episodes (25%). For manic episodes, EWS in thoughts racing, being full of ideas, and feeling agitated showed the highest sensitivity and specificity, whereas for depression, only feeling tired showed high sensitivity and specify.

EWS show promise in anticipating transitions to mood episodes in bipolar disorder. Further investigation is warranted.

No significant relationships.

Transitions in mental health, such as the onset or sudden progression of psychopathology, are difficult to foresee. If mental health behaves like other complex systems, drops in mental health may be anticipated by early warning signals (EWS), which manifest in the dynamics of time series data.

This study aimed to establish the sensitivity and specificity of EWS as personalized risk markers for sudden drops mental health.

Individuals (N=122, mean age 23.6 ±0.7 years, 57% males) at increased risk for psychopathology completed daily questionnaires on mental states for six consecutive months. Transitions in mental health were identified by change point analyses. EWS, operationalized as rising trends in the autoregressive coefficient of 36 negative mental states, were identified using generalized additive models.

EWS were found for 59% of individuals with a drop in mental health, and for 47% without such a drop (sensitivity: 0-.12; specificity: .88-1). There were considerable individual differences in the prevalence, strength, and timing of EWS.

EWS might be informative of impeding transitions, yet they are also highly conservative. Present findings may inspire future research into the prerequisites for detecting EWS in the context of mental health, for instance with respect to the stability of pre- and post-transition phases, the magnitude of transitions, and the timescale at which EWS manifest. An improved understanding of the dynamics that govern psychopathology could ultimately allow us to determine whether a specific individual at a specific moment in time is at risk for a sudden onset or progression of mental health problems.