Episode forecasting in bipolar disorder: Is energy better than mood?

OBJECTIVE: Bipolar disorder is a severe mood disorder characterized by alternating episodes of mania and depression. Several interventions have been developed to decrease high admission rates and high suicides rates associated with the illness, including psychoeducation and early episode detection, with mixed results. More recently, machine learning approaches have been used to aid clinical diagnosis or to detect a particular clinical state; however, contradictory results arise from confusion around which of the several automatically generated data are the most contributory and useful to detect a particular clinical state. Our aim for this study was to apply machine learning techniques and nonlinear analyses to a physiological time series dataset in order to find the best predictor for forecasting episodes in mood disorders. METHODS: We employed three different techniques: entropy calculations and two different machine learning approaches (genetic programming and Markov Brains as classifiers) to determine whether mood, energy or sleep was the best predictor to forecast a mood episode in a physiological time series. RESULTS: Evening energy was the best predictor for both manic and depressive episodes in each of the three aforementioned techniques. This suggests that energy might be a better predictor than mood for forecasting mood episodes in bipolar disorder and that these particular machine learning approaches are valuable tools to be used clinically. CONCLUSIONS: Energy should be considered as an important factor for episode prediction. Machine learning approaches provide better tools to forecast episodes and to increase our understanding of the processes that underlie mood regulation.

One-Shot Decoupling and Page Curves from a Dynamical Model for Black Hole Evaporation.

One-shot decoupling is a powerful primitive in quantum information theory and was hypothesized to play a role in the black hole information paradox. We study black hole dynamics modeled by a trilinear Hamiltonian whose semiclassical limit gives rise to Hawking radiation. An explicit numerical calculation of the discretized path integral of the S matrix shows that decoupling is exact in the continuous limit, implying that quantum information is perfectly transferred from the black hole to radiation. A striking consequence of decoupling is the emergence of an output radiation entropy profile that follows Page's prediction. We argue that information transfer and the emergence of Page curves is a robust feature of any multilinear interaction Hamiltonian with a bounded spectrum.

Exponential state transition dynamics in the rest-activity architecture of patients with bipolar disorder.

OBJECTIVE: Our goal was to model the temporal dynamics of sleep-wake transitions, represented by transitions between rest and activity obtained from actigraphic data, in patients with bipolar disorder using a probabilistic state transition approach. METHODS: We collected actigraphic data for 14 days from 20 euthymic patients with bipolar disorder, who had been characterized clinically, demographically, and with respect to their circadian preferences (chronotype). We processed each activity record to generate a series of transitions in both directions between the states of rest (R) and activity (A) and plotted the estimated transition probabilities (pRA and pAR). Each 24-hour period was also divided into a rest phase consisting of the eight consecutive least active hours in each day and an active phase consisting of the 16 consecutive most active hours in each day. We then calculated separate transition probabilities for each of these phases for each participant. We subsequently modeled the rest phase data to find the best fit for rest-activity transitions using maximum likelihood estimation. We also examined the association of transition probabilities with clinical and demographic variables. RESULTS: The best-fit model for rest-activity transitions during the rest phase was a mixture (bimodal) of exponential functions. Of those patients with rapid cycling, 75% had an evening-type chronotype. Patients with bipolar II disorder taking antidepressants had a lower probability of transitioning back to rest than those not on antidepressants [mean ± SD = 0.050 ± 0.006 versus 0.141 ± 0.058, F(1,15) = 3.40, p < 0.05]. CONCLUSIONS: The dynamics of transitions between rest and activity in bipolar disorder can be accounted for by a mixture (bimodal) of exponential functions. Patients taking antidepressants had a reduced probability of sustaining and returning to sleep.

Nonlinear dynamics of mood regulation in bipolar disorder.

OBJECTIVES: We sought to study the underlying dynamic processes involved in mood regulation in subjects with bipolar disorder and healthy control subjects using time-series analysis and to then analyze the relation between anxiety and mood using cross-correlation techniques. METHODS: We recruited 30 healthy controls and 30 euthymic patients with bipolar disorder. Participants rated their mood, anxiety, and energy levels using a paper-based visual analog scale; and they also recorded their sleep and any life events. Information on these variables was provided over a three-month period on a daily basis, twice per day. We analyzed the data using Box-Jenkins time series analysis to obtain information on the autocorrelation of the series (for mood) and cross-correlation (mood and anxiety series). RESULTS: Throughout the study, we analyzed 10,170 data points. Self-ratings for mood, anxiety, and energy were normally distributed in both groups. Autocorrelation functions for mood in both groups were governed by the autoregressive integrated moving average (ARIMA) (1,1,0) model, which means that current values in the series were related to one previous point only. We also found a negative cross-correlation between mood and anxiety. CONCLUSIONS: Mood can be considered a memory stochastic process; it is a flexible, dynamic process that has a 'short memory' both in healthy controls and euthymic patients with bipolar disorder. This process may be quite different in untreated patients or in those acutely ill. Our results suggest that nonlinear measures can be applied to the study of mood disorders.

Increased sympathetic tone is associated with illness burden in bipolar disorder.

BACKGROUND: We recently described an association between reduced heart rate variability (HRV) and illness burden in bipolar disorder (BD) using a novel Illness Burden Index (IBI). We aimed to further characterize this association by using spectral analyses to assess whether the IBI is also associated with autonomic imbalance in BD patients. METHODS: In this cross-sectional study, 53 participants with BD wore a device for 24 h to assess association between HRV spectral measures and the IBI or each of its components (age of onset, number and type of previous episode(s), duration of the most severe episode, history of suicide attempts or psychotic symptoms during episodes, co-morbid psychiatric disorders, and family history). We ran both unadjusted models and models controlling for age, sex, years of education, marital status, BMI, pharmacotherapy, and baseline functional cardiovascular capacity. RESULTS: HRV low-frequency (LF) normalized values were almost twice as high as published in healthy controls. Higher IBI was associated with higher LF and lower High Frequency (HF) values, resulting in a higher LF/HF ratio, indicating an increased sympathetic tone. Four individual components of the IBI were similarly associated with measures of increased sympathetic tone: earlier age of onset, number of depressive episodes, co-morbid anxiety disorders, and family history of suicide. Adjusted and unadjusted models had similar results. LIMITATIONS: Our models used mean LF and HF and do not consider their dynamic variations over 24 h or phase of the illness. CONCLUSIONS: Burden of illness is associated with increased sympathetic tone in patients with BD, putting them at risk for arrythmias and sudden death.

An admixture analysis of the age at index episodes in bipolar disorder.

The interaction between polarity at onset (PAO) and age at onset (AAO) appears to be important for interpreting results of previous analyses of AAO in bipolar disorder (BD). Using an admixture analysis, we examined independently the distributions of age at first depressive and hypomanic/manic episodes in 379 BD I and II patients. Subsequently, we examined the association of PAO and AAO with specific clinical variables, using parametric and nonparametric analyses. Both depressive and manic onsets showed bimodal distributions. For depressive episodes, the means were: 18.5±4.1 (early onset) and 33.6±10.4 (late onset) years; and for manic episodes 18.9±3.3 (early onset) and 34.8±10.9 (late onset) years. For the overall AAO the best fit was for a mixture of three lognormal distributions (mean±S.D.): 15.5±2.0, 22.8±4.6, and 36.1±10.1years. Overall, an early onset was significantly associated with a chronic course of the disorder, a stronger family history of affective disorder, higher rates of rapid cycling, suicidal behavior, psychotic symptoms, and co-morbid anxiety disorders. Early onset depressive episodes were associated with higher rates of suicidal behavior and anxiety disorders, whereas early onset manic episodes were associated with psychotic symptoms and rapid cycling. Our results suggest the presence of a bimodal distribution of age at onset in BD according to the polarity of the index episode, and denote that an early onset BD, irrespective of polarity, may be a more serious subtype of the disorder.

Reduced heart rate variability is associated with higher illness burden in bipolar disorder.

BACKGROUND: Bipolar disorder (BD) is associated with premature death and ischemic heart disease is the main cause of excess mortality. Heart rate variability (HRV) predicts mortality in patients with or without cardiovascular disease. While several studies have analyzed the association between HRV and BD, none has analyzed the association of HRV with illness burden in BD. METHODS: 53 participants with BD I and II used a wearable device to assess the association between HRV and factors characterizing illness burden, including illness duration, number and type of previous episode(s), duration of the most severe episode, history of suicide attempts or psychotic symptoms during episodes, and co-morbid psychiatric disorders. We ran unadjusted models and models controlling statistically for age, sex, pharmacotherapy, baseline functional cardiovascular capacity, BMI, years of education, and marital status. We also explored the association between HRV and an overall illness burden index (IBI) integrating all these factors using a weighted geometric mean. RESULTS: Adjusted and unadjusted models had similar results. Longer illness duration, higher number of depressive episodes, longer duration of most severe manic/hypomanic episode, co-morbid anxiety disorders, and family history of suicide were associated with reduced HRV, as was bipolar depression severity in the participants experiencing a depressive episode. Finally, a higher IBI score was associated with lower HRV. CONCLUSIONS: High illness burden is associated with reduced HRV in BD. While the IBI needs to be validated in a larger sample, it may provide an overall measure that captures illness burden in BD.

The futility of long-term predictions in bipolar disorder: mood fluctuations are the result of deterministic chaotic processes.

BACKGROUND: Understanding the underlying architecture of mood regulation in bipolar disorder (BD) is important, as we are starting to conceptualize BD as a more complex disorder than one of recurring manic or depressive episodes. Nonlinear techniques are employed to understand and model the behavior of complex systems. Our aim was to assess the underlying nonlinear properties that account for mood and energy fluctuations in patients with BD; and to compare whether these processes were different in healthy controls (HC) and unaffected first-degree relatives (FDR). We used three different nonlinear techniques: Lyapunov exponent, detrended fluctuation analysis and fractal dimension to assess the underlying behavior of mood and energy fluctuations in all groups; and subsequently to assess whether these arise from different processes in each of these groups. RESULTS: There was a positive, short-term autocorrelation for both mood and energy series in all three groups. In the mood series, the largest Lyapunov exponent was found in HC (1.84), compared to BD (1.63) and FDR (1.71) groups [F (2, 87) = 8.42, p < 0.005]. A post-hoc Tukey test showed that Lyapunov exponent in HC was significantly higher than both the BD (p = 0.003) and FDR groups (p = 0.03). Similarly, in the energy series, the largest Lyapunov exponent was found in HC (1.85), compared to BD (1.76) and FDR (1.67) [F (2, 87) = 11.02; p < 0.005]. There were no significant differences between groups for the detrended fluctuation analysis or fractal dimension. CONCLUSIONS: The underlying nature of mood variability is in keeping with that of a chaotic system, which means that fluctuations are generated by deterministic nonlinear process(es) in HC, BD, and FDR. The value of this complex modeling lies in analyzing the nature of the processes involved in mood regulation. It also suggests that the window for episode prediction in BD will be inevitably short.