Benzodiazepine use for anxiety disorders is associated with increased long-term risk of mood and substance use disorders: A large-scale retrospective cohort study

Abstract

Background: Benzodiazepines (BZDs) are widely prescribed for anxiety disorders. However, the long-term implications on mental health remain uncertain, especially the potential association between chronic BZD use and subsequent diagnosis of mood and substance use disorders (SUDs).

Method: We conducted a 5-year retrospective cohort study by analyzing the TriNetX database, a real-time electronic medical record network. The study population was defined as patients aged 18-65 with anxiety disorders (ICD-10-CM: F40-F48). We employed propensity score matching to pair a BZD-exposed cohort (≥12 BZD prescriptions) with a BZD-unexposed control cohort. The outcomes were defined as depressive disorders, bipolar disorders, and SUDs. We employed Kaplan-Meier analyses to assess the survival probability over five years following diagnosis and BZD exposure; log-rank test to obtain the hazard ratio (HR) with 95 % confidence interval (CI).

Results: We identified and matched 76,137 patients in the study and control cohorts. Compared to the control cohort, the BZD-exposed group exhibited significantly higher risks of being diagnosed with depressive disorders (HR, 2.64; 95 % CI, 2.59-2.68), bipolar disorders (HR, 4.39; 95 % CI, 4.15-4.64), overall substance use disorders (HR, 3.00; 95 % CI, 2.92-3.08), alcohol use disorder (HR, 3.38; 95 % CI, 3.20-3.57), stimulant use disorder (HR, 3.24; 95 % CI, 2.95, 3.55), cannabis use disorder (HR, 2.93; 95 % CI, 2.75-3.11), inhalant use disorder (HR, 4.14; 95 % CI, 3.38-5.06), and nicotine use disorder (HR, 2.72; 95 % CI, 2.63-2.81).

Conclusion: Our findings demonstrate a concerning association between BZD use and an increased risk of being diagnosed with various mood disorders and SUDs.

Keywords: Anxiety disorders; Benzodiazepines; Mood disorders; Substance use disorders.

Fig. 1

Hazard ratio significance of depressive disorders, bipolar disorders, and substance use disorders difference in the BZD cohort versus the control cohort. Depressive disorders were defined as diagnoses including depressive episode (F32), major depressive disorder, recurrent (F33), and dysthymic disorder (F34.1). Bipolar disorders were defined as diagnoses including manic episode (F30), bipolar disorder (F31), and cyclothymic disorder (F34.0). Substance use disorders were defined as mental and behavioral disorders due to psychoactive substance use (F10-F19). Base rates of participants developing each disorder were provided in supplement 1.