Is internet-based cognitive behavioral therapy for alcohol use disorder equally effective for men and women? Implications of a secondary analysis of a clinical trial.

Introduction: Excessive alcohol use is a major public health concern, for which internet interventions have shown to be effective. Group-average effects may however mask substantial inter-individual variations in changes; identifying predictors of this variation remains an important research question. Biological sex is associated with pharmacokinetic differences in alcohol tolerance, which is reflected in many national guidelines recommending sex-specific thresholds for excessive drinking. Whether effects of internet interventions are moderated by sex, and whether any moderation is due to confounders, remains largely unexplored. Aim: To examine sex-differences in outcomes (both response and remission) after an internet intervention for alcohol use disorder, and to identify any confounders. Method: The current study is a secondary analysis of a randomized controlled trial. After identifying factors in which men and women differed at baseline, mixed effects models were re-run using a subsampling matching strategy. Results: Men and women differed in baseline sum of drinks and self-rated anxiety. Sex was found to moderate (absolute) response but not remission, neither when using sex-specific or common thresholds for risky drinking. However, after controlling for baseline drinking through subsampling, the difference in response was no longer significant. Conclusion: Our findings suggest that the apparent sex-difference in treatment response was confounded by intercept-slope correlation - i.e. since men on average drank more at baseline, this offered larger room for decreasing. When conducting studies on internet interventions for addictive disorders, it is crucial to consider which outcomes to use, and how these are operationalized.