Supplementary Material for: Glucose intake reduces craving in patients with alcohol use disorder depending on insulin response

Introduction: There is evidence that the appetite-regulating hormone insulin plays an important role in alcohol use disorder (AUD), in the sense that there is a negative correlation between insulin and alcohol craving, meaning that an increase in insulin levels leads to a reduction in acute craving. This suggests a promising approach for the acute reduction of craving in the treatment of patients with AUD, which could be achieved via an actively induced short-term increase in insulin levels, e.g. by glucose administration, and which has not yet been investigated in the form of a randomized controlled trial. Another aspect that has not yet been investigated is the role of the insulin-responder-type of each individual, i.e. the time until the insulin peak is reached, in this relationship. Methods: The randomized, placebo-controlled, double-blind crossover study examined a glucose intake as acute treatment to reduce craving in 80 male and female patients with AUD. Dynamics in craving and insulin levels were assessed at eight time points on each study visit before and after alcohol cue exposure, after treatment with glucose respective placebo solution and during the subsequent observation phase. These changes were analyzed using linear mixed models. The insulin-responder-type (fast, normal, slow) of each person was taken into account and possible interactions with the treatment were analyzed. Results: Linear mixed models revealed a significant interaction effect (F(2,412.058)=7.988, p<.001) between treatment and insulin-responder-type on craving, with lower craving values in the glucose compared to the placebo condition in the normal insulin-responder-type group (i.e. insulin peak after glucose intake within 30 to 60 minutes, difference in means=-.805, p=.003, 95%CI: -1.428, -.182). In the contrary, in the fast insulin-responder-type group craving values were higher in the glucose compared to the placebo condition (difference in means=1.143, p=.011, 95%CI: .378, 1.907). Slow insulin responders showed no differences in craving levels depending on the treatment condition (difference in means=-.124, p=.694, 95%CI: -.741, .493). No main effect was found in the linear mixed models for baseline-centered insulin levels (F(1,395.337)=2.328, p=.128). Conclusions: Glucose intake may reduce craving in individuals with AUD who show a normal insulin response. Further research should consider the different insulin-responder-types and peak times to better understand the underlying mechanisms of craving reduction with glucose administration in the context of insulin elevation.