A research team led by 2022 BBRF Young Investigator Sarah H. Sperry, Ph.D., of the University of Michigan, has published results of a study of 481 people diagnosed with bipolar disorder (BD). They underline the importance of “mood instability” occurring between the major mood episodes that have traditionally provided a foundation for describing and treating the illness.

There are three diagnostic categories for bipolar disorder indicated in the DSM-V manual used by most psychiatrists. Those diagnosed with Bipolar I have one or more depressive and manic major mood episodes; those with Bipolar II have one or more episodes of depression and hypomania, the latter a less intense but equally important form of mania. A third category, BD NOS (Not Otherwise Specified) is characterized by mood episodes that don’t meet criteria in BD I or BD II for intensity or duration.

This diagnostic schema has provided “the foundation of BD research and clinical care,” Dr. Sperry and colleagues note in their new paper, “and has focused on characterizing and treating episodes [of depression or mania/hypomania], identifying predictors of episode relapse, and the short- and long-term consequences of episodes.”

However, the researchers point out, “increasing evidence suggests that bipolar disorders are associated with mood instability, even outside the context of [major] mood episodes.”

Dr. Sperry, in part with the support from her Young Investigator grant, has for a number of years devoted her research to developing and testing mood instability measures that might reveal the extent and importance of between-episode mood fluctuations. The new paper, which appears in Nature Mental Health, reports results that led the team to identify three subgroups of BD patients based specifically on mood instability. Each of the subgroups has potential predictive power regarding long-term outcomes and implications for how patients are assessed and treated, the investigators said. The team also included: Melvin G. McInnis, M.D., 1999 BBRF Independent Investigator and 1992 Young Investigator; and Ivy F. Tso, Ph.D., 2018 BBRF Young Investigator.

The team defines mood instability as “frequent and/or intense fluctuations in mood over time.” While mood fluctuations “inherently underlie the [existing] diagnostic criteria of BD,” they stress that mood instability “also includes shifts in mood symptoms at subsyndromal levels.” By this, they mean that some mood shifts may not reach the duration or intensity levels that qualify them as “major” episodes, but may nevertheless have a very real impact on the experience that patients have with the illness, and on how they fare over time. Among other things, the attention to “subsyndromal” mood shifts between major episodes calls into question the assumption made by some that in between major episodes, most BD patients are “euthymic,” i.e., living in a state of emotional stability and well-being.

The team made use of data from one of the largest long-term studies of BD, the Prechter Longitudinal Study of Bipolar Disorder (PLS-BD). They aimed to characterize and identify subgroups among 481 individuals based on mood instability measured by standard self-report-based instruments every 2 months over 5 years. They also used machine learning to identify various risk factors noted at baseline (when each participant joined the study) that might predict each participant’s mood instability status and their clinical and life-functioning outcomes in the 6th year after recruitment.

The cohort studied, mean age 41, was largely White (90%) and female (72%). 71% had a BD I diagnosis; 22% BD II; and 7% BD NOS.

Analysis revealed that participants fell into three distinct subgroups of mood instability, “high,” “moderate,” and “low.” Over 78% were in the high or moderate groups, suggesting to the team that mood instability is a non-trivial factor in understanding the patient experience.

Of the many potential risk factors identifiable at “baseline,” the team found seven that most reliably “predicted” outcomes (in retrospect, from the perspective of the investigators, who could not know at the outset which, if any, would have any predictive value 6 years later). The seven factors, in order of descending predictive power, were: neuroticism; sleep quality; childhood emotional neglect or physical abuse; stimulant abuse; age when hypomania first occurred; and number of depressive episodes. (The leading predictive factor, neuroticism, is used by psychiatrists to describe individuals with a tendency to experience negative emotions, such as anxiety, depression, anger, and self-doubt.)

“All seven of these factors can easily be measured across development and treatments, highlighting their utility as early risk markers,” the team wrote.

They also found significant differences in clinical and functional outcomes across the three mood instability subgroups, with high mood instability correlating with the most severe outcomes in year 6. Compared with those in the low instability class, those in the high instability class had worse family and work functioning, poorer mental health functioning, and higher suicidal ideation in year 6. Being in the high instability group also predicted worse family functioning and suicidal ideation compared with those in the moderate group. Being in the moderate group, in turn, predicted worse family functioning compared with those in the low instability group. Also, the results indicated that the more chronic or remitting that depression is in BD patients, “the more unstable mood tends to be going forward.”

“These results highlight the central role of mood instability in BDs, regardless of subtype or symptom severity,” the researchers concluded. Hence, in their view, “the present findings compel us to rethink how we assess and treat BDs.”

Dr. Sperry and colleagues say that that their results make the case for comprehensive and ongoing assessment strategies in clinical care of BD patients, as well as in longitudinal research of the illness. “We argue that baseline psychodiagnostics assessments should incorporate measures of the seven key predictors of mood instability identified in this study.” Doing so, they said, might facilitate early identification of mood instability risk even before diagnostic status (such as BD I or II) is established.

The team also argues in the paper for implementing patient-reported mood assessment measures, which could potentially reveal significant—and treatable—mood instability between major episodes.

The investigators indicated a range of goals for follow-up studies, which should, among other things, include a more diverse patient population, and establish precisely which assessment tools should be used, and how often and for how long to identify mood instability risk in individual cases. Ultimately, an optimized way of making such determinations “may enable earlier, targeted interventions that mitigate mood destabilization and improve long-term outcomes” in those with BD, they said.