Oxidative and nonoxidative benzodiazepines and the risk of femur fracture

Antonio Sgadari, Università Cattolica del Sacro Cuore, Campus di Roma
Kate L. Lapane, Brown University
Vincent Mor, Brown University
Francesco Landi, Università Cattolica del Sacro Cuore, Campus di Roma
Roberto Bernabei, Università Cattolica del Sacro Cuore, Campus di Roma
Giovanni Gambassi, Università Cattolica del Sacro Cuore, Campus di Roma
C. Gatsonis
L. Lipsitz
J. Hiris
C. Brostrup-Jensen
S. Gonzales
J. Hogan
O. Intrator
M. Barbour
K. Berg
E. Broderick
D. Forman
D. Gifford
A. Hume
P. Larratt
K. Steel

Abstract

Benzodiazepine use is a well-identified risk factor for falls and the resulting femur fractures in elderly adults. Benzodiazepines not requiring hepatic biotransformation may be safer than agents undergoing oxidation because oxidative activity has been shown to decline with age. The association between the use of either oxidative or nonoxidative benzodiazepines and the risk of femur fracture among elderly adults living in nursing homes was studied. A nested case-control study was conducted using the Systematic Assessment of Geriatric drug use via Epidemiology (SAGE) database; the records of 9,752 patients hospitalized for incident femur fracture during the period 1992 to 1996 were extracted, matching by age, gender, state, and index date to the records of 38,564 control patients. Conditional logistic regression models were conducted to estimate the odds ratios (ORs) for femur fracture with adjustment for potential confounders. The adjusted OR for the overall use of benzodiazepines was 1.10 (95% confidence interval [CI], 0.98-1.20); the risk seemed of only slightly greater magnitude for exposure to nonoxidative agents (1.18; 95% CI, 1.03- 1.36) than to oxidative benzodiazepines (1.08; 95% CI, 0.95-1.23). Among the latter, the effect was mainly accounted for by the use of agents with a long elimination half-life. A dose relationship was observed exclusively among users of long half-life oxidative benzodiazepines. The risk associated with the use of nonoxidative benzodiazepines showed no relationship to the age of the patients. In contrast, patients aged 85 years or older receiving oxidative benzodiazepines at high dosages or as needed had a two- to three- fold increased risk of femur fracture than did patients in the younger age group. Among older individuals, the use of benzodiazepines slightly increased the risk of femur fracture, mainly irrespective of the metabolic fate of the drug. Our results suggest that the use of nonoxidative benzodiazepines does not carry a lower risk for femur fracture than does the use of oxidative benzodiazepines. However, the latter agents may be associated with a somewhat higher risk of side effects among the oldest old, especially at higher dosages.