Abstract
To examine the mediation effect of burnout on the association between workaholism and tobacco and alcohol use. A total of 2199 workers from the French national electricity company fulfilled an online questionnaire. Smoking status, alcohol use disorder based on the Alcohol Use Disorders Identification Test-Consumption and workaholism based on the Work Addiction Risk Test were used as binary variables. Burnout was assessed as a continuous variable with the Copenhagen Burn-Out Inventory. Mediation analyses tested the direct effect of the associations between workaholism and each substance use, as well as the indirect effect passing through burnout, while adjusting for sociodemographic factors (gender, age, occupational grade and marital life), work stress using the effort-reward imbalance and overcommitment. When testing the mediation effect of burnout on the relation between workaholism and smoking, there was a significant direct effect of workaholism on smoking (Estimated effect of 0.27 [95% CI 0.01; 0.54]) and a significant indirect effect passing through burnout (Estimated effect of 0.09 [95% CI 0.02; 0.15]). When testing the mediation effect of burnout on the relation between workaholism and alcohol use, the direct effect of workaholism on alcohol use was not significant (Estimated effect of 0.21 [95% CI − 0.01; 0.44]) while the indirect effect passing through burnout was significant (Estimated effect of 0.10 [95% CI 0.04; 0.17]). Information and prevention regarding substance use should be reinforced among workers exposed to workaholism, especially if their workaholism led to a high level of burnout. Preventing the emergence of burnout among workaholics might have some benefits on their tobacco and alcohol use.