Benzodiazepine Prescription for Anxiety Disorders Increase the Risk of Substance Use Disorders: A Retrospective Cohort Study

Abstract

Introduction While the role of benzodiazepines (BZDs) has been well established for anxiety and related disorders, there are significant concerns about BZD dependence, withdrawal, and tolerance. There is a lot of ambiguity regarding the potential long-term effects of BZDs on mental health. However, the risk of developing subsequent other substance use disorders is in question. Objectives In this electronic medical record (EMR) based retrospective cohort study, the study cohort was defined as patients between the ages of 18 and 65 with anxiety disorders (ICD-10-CM: F40-F48) prescribed with at least one BZD; the control cohort was defined as patients between the ages of 18 and 65 with anxiety disorders (ICD-10-CM: F40-F48) with no BZD prescription during the five-year timeframe examined. We excluded patients with pre-existing substance use disorders (ICD-10-CM: F10-F19), et al. Methods We collected data from TriNetX Research database, a real-time international EMR network, from September 2017 to September 2022. Patients in the two cohorts were matched by gender, age, race, ethnicity, and common medical conditions at a 1:1 ratio by propensity scoring and then underwent Kaplan–Meier analysis and association analysis. Results A total of 626,754 patients were identified and matched for analysis. Patients in the study cohort were more likely to be female (67.6% vs. 66.7%, p < 0.001), non-Hispanic (65.8% vs. 62.5%, p < 0.001) and white (72.8% vs. 69.1%, p < 0.001). Kaplan–Meier analysis showed the survival probability at the end of the time window was 94.1% for the control cohort and 89.5% for the study cohort (Hazard ratio, 2.20; 95% CI, 2.16-2.25; P < 0.001) in all type of substance use disorders. (Table 1) Table 1. Hazard ratio of substance use disorders difference in BZD cohort versus the control cohort. BZD Cohortn (risk%) Control Cohortn (risk%) Hazard Ratio(95% Cl) P value Substance Use Disorders* 26,569 (4.2) 11,976 (1.9) 2.20 (2.16- 2.25) <0.001 Sedative/hypnotic/anxiolytic related disorders 656 (0.1) 152 (0.0) 4.26 (3.57- 5.09) <0.001 Alcohol Related Disorder 5,749 (0.9) 2,064 (0.3) 2.74 (2.61-2.88) <0.001 Opioid Related Disorder 2,807 (0.4) 815 (0.1) 3.38 (3.13-3.66) <0.001 Stimulant Related Disorder 1,658 (0.3) 551 (0.1) 2.94 (2.67- 3.24) <0.001 Cannabis Related Disorder 3,376 (0.5) 970 (0.2) 3.41 (3.17- 3.66) <0.001 * Substance use disorders was defined as Mental and behavioral disorders due to psychoactive substance use (ICD-10-CM: F10-F19). Conclusions Patients with an anxiety disorder who were prescribed BZDs are at higher risk of not only BZD dependence but all types of substance use disorders than a matched cohort not prescribed BZDs. Given this notable association, clinicians should be cautious while prescribing BZDs and inform the patient about the risks associated with their utilization. Disclosure of Interest None Declared