Background
Daily variation in mood and activity between and within affective episodes in bipolar disorder has become a field of increasing interest. The present exploratory study aimed to identify predictors of mood and activity instability in participants with bipolar disorder (BD).

Methods
A total of 258 participants with newly diagnosed BD type I and II were included as part of a longitudinal study (the Bipolar Illness Onset study (BIO)). The participants completed daily smartphone-based mood and activity ratings for a median [interquartile range] (IQR) of 109 days [40;240] and 106 days [40;243], respectively. Clinical evaluations and questionnaires were collected at the baseline visit. Backwards stepwise regression analysis was employed to identify predictors.

Results
Predictors of increased mood instability included childhood trauma (e.g., B = 0.006, 95 % CI 0.001;0.011, p = 0.031), increasing number of depressive episodes, antipsychotic medication, functional impairment, and impaired sleep quality. Predictors of increased activity instability included longer Illness duration, lower age at onset, number of depressive episodes, functional impairment, and impaired sleep quality (e.g., B = 0.089, 95 %CI=0.033;0.145, p = 0.002).

Limitations
Risk of type 1 error due to the large number of analyses.

Conclusions
Increasing number of prior depressive episodes, functional impairment, and poor sleep quality were consistent predictors of subsequent increased mood and activity instability. Childhood trauma was a predictor of increased mood instability only, whereas age of onset was a predictor of increased activity instability only.