Background:
An estimated 6% – 10% of U.S. adults took a hypnotic drug for poor sleep in 2010. At least 18 studies have reported significant (p<0.05) associations of hypnotic usage with increased mortality. However, most lacked data on newer, supposedly safer, short-acting drugs, and had limited control for confounding by health status. Furthermore, little is known regarding potentially heightened risks in specific vulnerable populations.
Objective:
The present study was designed to test whether newer short-acting hypnotic drugs were associated with increased mortality after controlling for comorbid conditions, and to assess risks within subgroups of patients with specific medical conditions.
Methods:
Using electronic medical records from a large U.S. health system the authors conducted a one-to-two matched-cohort survival analysis of associations between hypnotic drug use and mortality. Records were extracted for 10,529 hypnotic users and 23,676 matched controls with no hypnotic prescriptions, mean age 54 years, followed for an average of 2.5 years between 2002 and 2006. Hazard ratios (HR) for death were computed from Cox models controlled for risk factors and stratified on comorbidites.
Results:
The short-acting drugs zolpidem (41%) and temazapam (20%) accounted for the majority of use. Patients prescribed any hypnotic had substantially elevated hazards of dying compared to non-users. Importantly, the death hazard was evident even in the lowest tertile, 1 to 18 pills per year, HR 3.60 (95% Confidence Interval, 2.92 – 4.44). HRs for the remaining tertiles were 4.43 (3.67 – 5.36) and 5.32 (4.50 – 6.30), demonstrating a dose-response association. HR were robust within subgroups restricted to users and non-users with identical comorbidity, implying that selective use of hypnotics by patients in poor health was an unlikely explanation for the excess mortality. Obesity emerged as a marker of increased vulnerability. Among 2206 patients with a diagnosis of obesity, (mean BMI 38.8), the mortality HRs by hypnotic tertile were 8.07 (3.64 – 17.89), 6.37 (2.73 – 14.88), and 9.34 (4.47 – 19.52). Additional models were fitted for patients with the combination of Obesity + Diabetes + Hypertension to evaluate the possibility that this risk was driven by metabolic syndrome. HRs for that combination were slightly lower than those for obesity alone, suggesting that obesity was the primary factor.
Conclusions:
Short-acting hypnotics were associated with a more than 3-fold increased hazard of death that, even at low levels of use. Obese patients appear particularly vulnerable, perhaps through interaction with sleep apnea. Emerging evidence for substantial harm, even with limited exposure to hypnotics, should be weighed against any benefits.
Use of hypnotic drugs in a learning disabilities hospital
