Background: The prevalence of substance use disorders in bipolar disorder (BD) is high. Exploring potential interactions between mood and the use of common substances such as alcohol and nicotine may contribute to a better understanding of the mechanisms underlying such comorbidities. Digital tools now allow for continuous monitoring and data collection of both symptoms and behavior. This enables time-series analyses to explore such associations with greater precision.
Methods: Thirty-two individuals in the early phases of BD registered their mood daily and their use of substances weekly in the MinDag (MyDay) app for up to 6 months. We explored temporal relationships between the use of alcohol and nicotine and the levels of depressed, elevated, irritable, and anxious mood using Vector Autoregressive Models and Granger causality tests.
Results: We found indications that mood influenced alcohol- and nicotine use, and vice versa. Significant temporal relationships (Granger causality) were found in 55% (11 out of 20) of the participants for alcohol and 70% (7 out of 10) for nicotine use, and with high proportions of the variance explained by the one time-series on the other. The associations were consistent with causal effects in one or both directions, but with no adjustment for confounding.
Conclusion: Our findings indicate that mood influences alcohol- and nicotine use and vice versa in individuals with BD, although caution should be taken due to the exploratory approach. Larger samples are needed to further disentangle these relationships to provide insight for better prevention and treatment of BD and comorbid substance use disorders.
Keywords: Alcohol use; App monitoring; Bipolar disorder; Mood; Nicotine use; Time-series analyses.
© 2025. The Author(s).