Preliminary animal and human studies have shown that blood dihydrotestosterone concentrations are increased in males with alcohol use disorder, and 5α-reductase inhibitors, which decrease dihydrotestosterone concentrations, reduce alcohol consumption. To gain mechanistic insight, we studied the effects of reduced dihydrotestosterone concentrations following pharmacological 5α-reductase inhibition on alcohol cue-elicited brain activity and alcohol craving in males with problematic alcohol use. To this end, this randomized, placebo-controlled, crossover challenge experiment investigated associations between dihydrotestosterone concentrations and brain functional magnetic resonance imaging (fMRI) activity during exposure to visual alcohol cues and alcohol craving following a single dose of 5 mg finasteride versus placebo in 50 males with heavy episodic drinking. We used finasteride because it specifically inhibits 5α-reductase II activity, which is the main enzyme converting testosterone to dihydrotestosterone. Dihydrotestosterone concentrations were lower in the finasteride condition in comparison to the placebo condition, but not significantly associated with brain activation patterns or craving. In the exploratory analyses, we found higher brain activity during exposure to visual stimuli in the right and left caudate nuclei, the right superior frontal gyrus and the left insula in the finasteride condition versus the placebo condition. Moreover, finasteride versus placebo was associated with a higher wish to not drink alcohol. The results of this experimental study do not support the à priori hypothesis that dihydrotestosterone concentrations play a role in brain activation during exposure to visual alcohol cues, but indicate that finasteride effects may be mediated by other pathways. Future studies are requested to investigate the effects of reduced dihydrotestosterone concentrations over a longer time and to shed light on the molecular mechanisms underlying the here observed effects of finasteride. Trial Registration: DRKS00020569.
Boroumand‐Jazi et al. (Wed,) studied this question.