Premium
A novel technique to evaluate fluctuations of mood: implications for evaluating course and treatment effects in bipolar/affective disorders
Author(s) -
Sree Hari Rao V,
Raghvendra Rao C,
Yeragani Vikram K
Publication year - 2006
Publication title -
bipolar disorders
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.285
H-Index - 129
eISSN - 1399-5618
pISSN - 1398-5647
DOI - 10.1111/j.1399-5618.2006.00374.x
Subject(s) - mania , mood , bipolar disorder , psychology , depression (economics) , mood disorders , psychiatry , clinical psychology , anxiety , economics , macroeconomics
Objectives: Several psychiatric conditions are associated with frequent fluctuations of affect. In this study, we propose a new technique to uniformly score depression and mania objectively and use a new mathematical technique to model the frequent fluctuations in mood using simulated data. Our main aim is to examine the usefulness of this measure for evaluating treatment effects or course of illness, especially in bipolar or unipolar affective illness to quantify mood fluctuations. Methods: We use a prototypical model, which takes into account the mean, the standard deviation (SD) and the coefficient of variation (CV = SD*100/mean) of the mood scores of the subjects over a user‐defined period. We utilize simulated data of subjects for euthymia, minor depression, minor mania, severe depression, severe mania and cyclic bipolar illness (manic depression, MDP). We propose an objective method to quantify the mood of the subjects at weekly intervals (the interval can be user‐defined) using a scale of 1–9 (1–4 = degrees of depression, 5 = euthymia, 6–9 = degrees of mania). These scores can be sampled according to the convenience and feasibility of the measurements, which can be derived from various clinical scales or by observation of the subjects in hospitals or other environments. We derive a new mathematical technique to arrive at a normalized measure for each of these conditions of simulated data in addition to the mean, SD and approximate entropy (ApEn). Results: We utilize three sets of data, one to train the model to classify the condition of the subjects and the other two to test the reliability of the technique. We are able to successfully classify the condition of the subjects over a 52‐timepoint period (length can be days or weeks depending upon the sampling rate). The New Index (NI) correlates significantly only with the mean ( r 2 = 0.78), but not with the SD or ApEn score. Conclusions: These results indicate that it may be beneficial to reduce data according to the techniques we propose so that there is greater uniformity within which to compare future studies to evaluate treatment effects, not only in rapid‐cycling MDP but also in other affective disorders. This method may be suitable for the meta‐analysis of several studies, although different scales have been used in each of those studies. Our measure derived from simulated data has shown sufficient deviation of all the abnormal states from the euthymic state. The advantages and pitfalls of these techniques are further discussed to evaluate affect in various disorders. However, future prospective studies must address the importance of this measure in comparison with mean, SD and ApEn scores or other nonlinear measures of these time series. We are evaluating other nonlinear dynamic models, which may provide a continuous measure with which to identify different degrees of fluctuation of mood.