Abstract
BACKGROUND: Evidence suggests that alcohol use and smoking are negatively associated with mood in bipolar disorders (BD). It is unknown if this relationship is moderated by the number of previous mood episodes. Therefore, this paper aims to examine whether the number of previous mood episodes moderates the relationship between alcohol use and smoking, and mood. METHOD: This study assessed the outcomes of 108 outpatients with BD I and II in a prospective observational cohort study. For 1 year, subjects daily registered mood symptoms and substance use with the prospective Life Chart Method. The relationship between the average daily consumption of alcohol and tobacco units in the whole year and mood were examined by multiple linear regression analyses. Number of previous mood episodes, grouped into its quartiles, was added as effect moderator. Outcome was the number of depressive, hypomanic and manic days in that year. RESULTS: The number of depressive days in a year increased by 4% (adjusted β per unit tobacco = 1.040; 95% CI 1.003-1.079; p = 0.033) per unit increase in average daily tobacco consumption in that same year. Interaction analyses showed that in those subjects with less than 7 previous mood episodes, the number of manic and hypomanic days increased by 100.3% per unit increase in alcohol consumption (adjusted β per unit alcohol = 2.003; 95% CI 1.225-3.274; p = 0.006). In those with 7 to 13 previous mood episodes, the number of manic and hypomanic days decreased by 28.7% per unit increase in alcohol consumption (adjusted β per unit alcohol = 0.713; 95% CI 0.539-0.944; p = 0.019); and in subjects with 14 to 44 previous mood episodes, the number of manic and hypomanic days decreased by 7.2% per unit increase in tobacco consumption (adjusted β per unit tobacco = 0.928; 95% CI 0.871-0.989; p = 0.021). CONCLUSIONS: The number of previous mood episodes moderates the relationship between alcohol use and smoking and mood; and smoking is adversely associated with the number of depressive days.