Prescription hypnotics and the risk of cardiovascular death

Vanderbilt University Medical Center, Nashville TN

Investigators

Linked publications & trials

Abstract

Hypnotic prescriptions for U.S. adults increased from 5.3 million in 1999 to 20.8 million in 2010, with the most frequent use by persons 65 years of age or older. The 3 most commonly prescribed hypnotics are the benzodiazepines, selective benzodiazepine receptor agonists (sBzRAs), and trazodone. Observational studies consistently have found that benzodiazepines and sBzRAs are associated with a 60 to 73% increased risk of all-cause mortality. Controversial biological plausibility and study limitations have led most to conclude that the association is not causal. However, given the large number of patients taking hypnotics, if even some part of the observed association were causal, there would be major public health consequences. One mechanism by which benzodiazepines and sBzRAs could increase mortality is adverse effects on nocturnal respiration. Nocturnal respiratory impairment causes apnea, hypoxemia and hypercapnia, which are associated with sympathetic activation and increased risk of cardiac arrhythmias and cardiovascular death. Benzodiazepines consistently impair respiration, sBzRAs do so to a lesser degree and the available data indicate trazodone causes minimal, if any, respiratory impairment. This suggests that trazodone could have better cardiovascular safety than benzodiazepines and possibly sBzRAs. Opioid analgesics frequently are co- prescribed with benzodiazepines and sBzRAs. Thus, synergistically respiratory impairment might potentiate the risk of cardiovascular death for these hypnotics. Thus, patients and practitioners urgently need reliable information on the relative cardiovascular safety of hypnotics, both without and with concurrent opioid use. Because the relevant randomized clinical trial is unlikely to be performed, the data must come from an epidemiologic study. To impact practice, the numerous limitations of previous observational studies must be avoided. We plan to address these questions by continuing our productive series of epidemiologic investigations and apply previously developed tools to study hypnotic safety. We will conduct a retrospective cohort study with a rigorous design in Medicare beneficiaries?with both the most frequent hypnotic use and greatest susceptibility to increased mortality. We will test two hypotheses: Aim 1. In the absence of opioid use, the risk of out-of-hospital cardiovascular and total mortality for benzodiazepine and sBzRA users is greater than that for trazodone users. Aim 2. Concurrent use of opioid analgesics potentiates the risk of cardiovascular mortality for patients with benzodiazepine or sBzRA use.