Panic disorder, alcohol and substance abuse, and benzodiazepine prescription.

Abstract

Sir: In a recent supplement of the Companion, the utilization of benzodiazepines in clinical practice was reviewed.1 In that supplement, several authors advised caution in prescribing benzodiazepines to patients with histories of alcohol (p. 5) or substance abuse (p. 23). We were curious about the actual percentage of patients for whom this precaution might apply. To explore this, we undertook a literature search dating back to 1990 using the PsycINFO and MEDLINE databases. We entered the search terms panic disorder, substance abuse, and alcohol. After identifying and gathering articles, we excluded those that had already been addressed in review articles, were written in foreign languages, or reported studies that comprised adolescent participants. We first examined the lifetime prevalence of alcohol and substance abuse in studies in which participants were identified by a diagnosis of panic disorder. As a prelude to our findings, we encountered a 1990 review article by Brady and Lydiard2 that reported 2 earlier studies in which the prevalence of alcohol abuse or alcoholism was 7% to 8% among panic disorder samples. We located 7 studies published since 1990 (Table 1) that explored the lifetime prevalence of alcohol and substance abuse in participants with panic disorder.3–9 Because the study by Lepine et al.4 did not separate alcohol from substance abuse, we elected to exclude it from the present analysis, bringing the subsequent working sample size to 844 participants. In this subsample, 150 participants (17.7%) with panic disorder were also diagnosed with alcohol abuse or dependence. As for “other substance abuse,” again excluding the Lepine et al. study,4 4 studies comprising 500 participants were relevant. The prevalence of other substance abuse was confirmed in 111 participants (22.2%). In summary, these data indicate that in studies since 1990, approximately 20% of panic disorder patients had lifetime histories of either alcohol or other substance abuse. Table 1. Lifetime Prevalence of Alcohol and Substance Abuse Among Study Participants With Panic Disorder To augment the preceding findings, we examined 2 large community studies that explored the prevalence of alcohol and substance abuse in panic disorder participants during the 12 months preceding assessment. In the first study,10 the prevalence rate of panic disorder among 9282 participants was 2.7%; the correlations between panic disorder and alcohol abuse, alcohol dependence, drug abuse, and drug dependence were 0.27 (significant at p < .05), 0.25, 0.16, and 0.27, respectively. In the second study,11 the prevalence of any comorbid alcohol use disorder during the preceding 12 months in panic disorder participants with and without agoraphobia was 18.8% and 15.3%, respectively, while the prevalence of any comorbid substance abuse disorder was 24.2% and 17.3%, respectively. We next examined the prevalence of panic disorder in study samples of alcoholics or substance abusers. Brady and Lydiard2 summarized these data up to 1993. Averaging various samples, these authors found that 8.8% (range, 1.0%–20.0%) of participants met the criteria for panic disorder. In the 2 subsequent studies12,13 that emerged during our literature search, the lifetime prevalence rate for panic disorder among alcoholics was 4.1% (144/3475), and among those who abuse or are dependent on sedative/hypnotics, the lifetime prevalence rate was 13.2% (58/441). These rates are substantially less than the rates of alcohol/substance abuse reported in panic disorder samples. Understandably, it is difficult to accurately assess and compare prevalence rates among studies because of varying assessment tools, different time frames (preceding 12 months vs. lifetime prevalence rates), the timing of the assessment in relationship to detoxification in alcohol and substance abuse populations, patient candor, and varying characteristics of the study population (e.g., primary vs. tertiary care, clinical vs. non-clinical). In addition, we may have missed studies presented at meetings and abstracts, as well as published articles that were not accessed by our search terms. However, these data suggest that approximately 20% of panic disorder patients have lifetime histories of alcohol and/or substance abuse. Because of the substantial rate of comorbidity between panic disorder and alcohol/ substance abuse, clinicians need to screen for these disorders at the outset and exercise caution in prescribing benzodiazepines in this subgroup other than for detoxification.