Prazosin for Alcohol Use Disorder: Reply to Sinha

See letter by Sinha

J Clin Psychiatry 2021;82(6):21lr14076a

To cite: Andrade C. Prazosin for alcohol use disorder: reply to Sinha. J Clin Psychiatry. 2021;82(6):21lr14076a.
To share: https://doi.org/10.4088/JCP.21lr14076a

© Copyright 2021 Physicians Postgraduate Press, Inc.

^aDepartment of Clinical Psychopharmacology and Neurotoxicology, National Institute of Mental Health and Neurosciences, Bangalore, India
*Corresponding author: Chittaranjan Andrade, MD, Department of Clinical Psychopharmacology and Neurotoxicology, National Institute of Mental Health and Neurosciences, Bangalore 560 029, India ([email protected]).

To the Editor: I do not contest the points that Dr Sinha makes.¹ However, I do wish to provide clarifications about the issues that she has raised. With regard to the first point, whereas statistical analyses are more appropriately performed on constructs that are operationalized as continuous variables, clinical decision-making often necessitates the examination of constructs as categorical variables.^2,3 This means that, in an analysis of continuous data, if a study finds that prazosin is associated with better alcohol use disorder (AUD) treatment outcomes when alcohol withdrawal symptom (AWS) ratings are higher, clinicians who treat AUD would want to know beyond what point AWS scores indicate potential benefits with the drug. In fact, in her letter, Dr Sinha herself acknowledges that “the premise was to assess whether an easy, clinically useful measure of AW may serve as a prognostic indicator of treatment to improve drinking outcomes.” Happily, the required information was available in the secondary analyses, and this information was presented to the reader in my commentary.⁴ It may be noted here that the purpose of my commentary was not to mechanically summarize the original study⁵ but to present to the reader theoretically and practically useful messages that emerged from the findings of the study.

With regard to the second point, a review of prazosin for the treatment of posttraumatic stress disorder nightmares identified sedation as an adverse effect of the drug.⁶ More specifically, previous RCTs of prazosin in patients with AUD,^7,8 which uptitrated prazosin in the same way that Sinha et al⁵ did, found significantly more drowsiness with prazosin than with placebo. If the RCT by Sinha et al⁵ failed to identify drowsiness or sedation with prazosin, it may have been because adverse effects were assessed weekly, or because the sedating effect may have been therapeutic for AUD-related sleep disturbances such as insomnia, and so was not elicited as an adverse effect.

With regard to the final point, regardless of the premise of the original study,⁵ it is the prerogative of a commentator to interpret findings and to construct hypotheses in the light of the background literature, which is exactly what I did.

On a parting note, Dr Sinha writes, “we did not find that prazosin reduced alcohol withdrawal symptoms over placebo.” However, in Table S4 in the online supplement to the study,⁵ the prazosin vs placebo treatment effect was statistically significant for the baseline Clinical Institute Withdrawal Assessment for Alcohol, Revised (CIWA-Ar) by week interaction, indicating (as stated in a footnote to the table) “reduction in CIWA-Ar scores across weeks but only in those with higher AW (cont.) scores and no change by week in those with no/low levels of AW.” So, it appears reasonable to speculate that the benefits of prazosin in patients with AUD emerge through reduction of AWS in persons with high baseline AWS scores.

Published online: September 21, 2021.
Potential conflicts of interest: Please see https://www.psychiatrist.com/jcp/about-dr-andrade.
Funding/support: None.