Research led by 2022 BBRF Young Investigator Sarah H. Sperry, Ph.D., of the University of Michigan, has led to her team’s development of an innovative method for assessing the course of bipolar disorder (BD) in those who have been diagnosed.

The new method, which focuses on assessing mood instability in individual patients—frequently and over an extended period—could result, if it were widely adopted, in what Dr. Sperry and colleagues call “a paradigm shift in monitoring outcomes” in BD.

Accurate diagnosis of bipolar disorder is often delayed, Dr. Sperry and colleagues note in their new paper appearing in Nature Mental Health, “and even when diagnosed properly, efficient treatment options remain stagnant.” It is standard practice, they explain, to think of BD as a “relapsing disorder,” in which distinct episodes of depression, mania, or hypomania alternate with periods of normal mood. The deep lows of depressive episodes stand in contrast to the extreme highs of mania in Bipolar I or the somewhat lesser highs of hypomania in Bipolar II.

“Relapsing” is an important term and point of reference for the research by Dr. Sperry and colleagues. As BD is usually thought of, in between depressive episodes on the one hand, and episodes of either mania or hypomania on the other hand, there are periods of “remission,” which are conceived as those periods of time when “euthymic” or normal mood prevails—neither qualifying as depressed nor manic/hypomanic. These periods can be of widely varying duration, long or short, depending on the patient.

But, say Dr. Sperry and colleagues, including senior author on the new paper, Melvin G. McInnis, M.D. (a 1999 BBRF Independent Investigator and 1992 Young Investigator), the picture is changing because, in recent years, more and more attention has been devoted to studying people with BD over time—what are called longitudinal studies—and assessing them repeatedly over short time periods. There are now wearable digital technologies that enable monitoring of certain behavioral patterns minute by minute and hour by hour. Assessments in the new study did not employ such technologies, but rather depended on patients’ self-assessments every 2 months.

In the study that gave rise to their new “paradigm” for assessing BD, Dr. Sperry and colleagues used data from 603 people collected over 10 years or longer for each individual. These people had been enrolled in the Prechter Longitudinal Study of Bipolar Disorder (PLS-BD), and included 385 people diagnosed with BD, 71 with other or nonaffective diagnoses, and 147 with no history of psychiatric diagnosis and none among first-degree relatives. The participants were typically in their 30s at study entry and White; about 60% were female.

What made this study distinctive was the repeated monitoring of participants, every 2 months, over 10 or more years. The so-called PROMS (patient-reported outcome measures) included assessment questionnaires for depression, mania/hypomania, anxiety, and overall mental and physical functioning.

The scores generated by all of these questionnaires were used to calculate the variation in each measure for each participant, combining results over rolling 1-year periods, as the study progressed. Advanced statistical calculations enabled the team to establish three broad “thresholds” for assessing the degree of variations in symptoms—labeled “mood (in)stability”—over the entire cohort. Low, moderate, and high instability were identified, based on comparisons of rolling 1-year individual data assessments across the different questionnaires.

The purpose in designing the study in this way was to test the team’s hypothesis about a new way of assessing what each patient with BD experiences over the course of time. Are “remissions” between “episodes” the best way to characterize what happens to most people with BD?

This question has great bearing on what kind of care patients receive. Current gold-standard clinical care of BD aims to reduce symptoms or achieve remission in both depressive and manic/hypomanic episodes, as well as to prevent such episodes from occurring where possible. Dr. Sperry has devoted her young career to investigating whether some or even many BD patients experience important mood shifts—significant “mood instability”—even in those periods usually thought of as “between” episodes. Evidence of such instability would call into question the standard assumption in clinical medicine that the “in-between” periods are ones of essentially “normal” mood, what psychiatrists call euthymia. If there is considerable mood instability “in between” episodes of diagnosable depression and mania/hypomania, how might that affect the care BD patients receive and their outcomes?

With the hypothesis that “alternative ways to measure change in BD [status] based on mood instability are needed,” the team did arrive at what they consider a “clinically meaningful instability score that is simple to calculate and easy to interpret.” They were also able to identify low, moderate, and high instability thresholds, based on individual scores, and used these classifiers to try to predict how individual patients fared both mentally and physically over time.

First, compared with participants in the study who had psychiatric issues other than BD as well as healthy controls, those participants with a BD diagnosis consistently had higher 1-year rolling “mood instability” scores. The effects were largest with respect to symptoms of depression and mania/hypomania, and smaller but still significant for anxiety. Those with BD, compared with those with other psychiatric issues, had larger instability variations in their scores within the 1-year time horizons, indicating to the team that this specifically reflects the trajectory of BD.

Regarding the power of the scores to predict: those participants with higher rolling 1-year instability scores in the measure of depression symptoms tended, if they were in the moderate- and high-threshold subgroups, to have worse mental and physical health functioning. Other questionnaire-based assessors of variations in individual mood status over rolling one-year periods were useful, in various ways, for predicting other outcomes for BD patients as the months and years of the study passed.

Broadly speaking, the team said, their results “provide guidelines for practical clinical monitoring in the daily patient-care setting, and offer an innovative strategy for outcomes assessments.” They said that their results argue for the clinical relevance of monitoring mood instability in individuals over time, specifically the ranges within which the various assessment tools fluctuate over time in individual patients.

Assuming that norms for mood instability within BD can be fully established and broadly recognized by the medical community, the team said, it might then be possible for research to tackle the question of how to reduce mood instability—fluctuations over comparatively short periods of time within each patient—as a way of improving their outcomes, which include not only mental and physical functioning, but also feelings of well-being, cognition, interpersonal relationships, and occupational outcomes.