Benzodiazepines Tied to a 41% Increased Mortality Risk in AD

Batya Swift Yasgur, MA, LSW

November 28, 2017

Patients with Alzheimer's disease (AD) who use benzodiazepines and related drugs (BZRDs) have a 41% higher risk for death than patients who do not use these drugs, new research shows.

Mortality rates in patients with AD who use BZDRs was 13.4 per 100 person-years, vs 8.5 per 100 person-years in nonusers during the 6-month study period (adjusted hazard ratio [HR], 1.4). The association was significant from initiation of use.

"I was surprised by how big the increased risk was," lead author Laura K. Saarelainen, a PhD candidate at the Kuopio Research Center for Geriatric Care, University of Eastern Finland, told Medscape Medical News.

"We would like clinicians to know that these drugs have major adverse events from the very beginning of use," she added.

The study was published online November 15 in the International Journal of Geriatric Psychiatry.

Vulnerable Population

Individuals with AD are a "vulnerable population," subject to a high frequency of chronic comorbidities, mortality, and the behavioral and psychological symptoms of dementia (BPSD), the authors write.

Although treatment guidelines recommend that BZDRs be used only as short-term or infrequent treatment of BPSDs, use of BZDRs "increases considerably" at the time of AD diagnosis, with almost one third of patients using these agents, the authors state.

Only a few studies to date have investigated adverse outcomes associated with BZDR use in this population. This previous research has found these drugs to be related to increased risk for hip fracture, stroke, and pneumonia.

Moreover, no previous studies have focused specifically on community-dwelling individuals with AD. The authors therefore aimed to study all-cause 180-day mortality in a nationwide cohort of community-dwelling persons with AD.

"We chose benzodiazepines for study because they have not received a great deal of previous research, and information about adverse effects is lacking," Saarelainen said.

To investigate this issue, the researchers used data from the Medication Use and Alzheimer's Disease cohort, which includes all community-dwelling persons diagnosed with mild to moderate AD in Finland between 2005 and 2011. Data were obtained from nationwide Finnish registers.

BZDRs were defined as benzodiazepines (anatomic therapeutic chemicals of classes N05BA and N05CD) and benzodiazepine-related drugs ("Z-drugs," or drugs of class N05CF).

Benzodiazepines included in the study were diazepam, chlordiazepoxide, oxazepam, lorazepam, alprazolam, nitrazepam, temazepam. Z-drugs included zopiclone and zolpidem.

Information on drug use included when each period of drug use started and ended. Regularity of drug purchases, hospital stays, and possible stockpiling of drugs were taken into account.

Patients were excluded if they had used BZDRs during a 1-year washout period prior to AD diagnosis, if they had been hospitalized or institutionalized for 50% or longer of the washout period, or if they had an ongoing hospital stay of 90 days or longer at the end of the washout period.

So long as use of these agents was continuous, overlapping BZDR and Z-drug use periods were combined. However, for subanalyses, periods of BZDR use and Z-drug use were analyzed separately.

Covariates included comorbidities, socioeconomic position, psychiatric disorders, substance abuse, stroke, antidepressant use, antipsychotic use, and opioid use.

After exclusion and inclusion criteria had been applied, the study included 10,380 individuals who initiated BZDR use and 20,760 matched nonusers. The study follow-up consisted of 3319.0 and 9282.3 person-years of BZDR use and nonuse, respectively (median age of participants, 81.1 years).

The median time from the date of AD diagnosis and the start of follow-up was 445 days (IQR, 166 - 903.5); median duration of BZDR use was 121 days (IQR, 40 - 180).

During the study period, 6438 patients initiated drug use with a BZDR, and 3826 initiated use with a Z-drug; 161 used both types of drugs concomitantly.

In total, 440 patients (3319.0 person-years) died during BZDR use, compared to 785 (9282.3) during nonuse. This translated into an age-adjusted death rate (per 100 person-years) of 13.4 (12.23 - 14.53) for the patients who used BZDRs, vs 8.5 (7.92 - 9.05) for those who did not use BZDRs (nonadjusted HR, 1.56; 95% CI, 1.38 - 1.77; adjusted HR, 1.41; 95% CI, 1.23 - 1.62).

The association between BZDR use and increased risk for death was significant but did not persist after 120 days. Moreover, Z-drug use was not associated with increased risk.

The writers note that their observation of a 41% increased risk for death with BZDR use was "higher than the observations in previous studies investigating persons with dementia."

Moreover, because the difference between increase in death in patients who used BZDRs and those who used Z-drugs was very small, "Z-drug use cannot be considered any safer than benzodiazepine use in persons with dementia regarding risk of death," the authors note.

"We believe that there is no difference between these two drug categories and that the lower risk associated with Z-drugs is that they are less frequently used," Saarelainen said.

The researchers attribute the adverse outcomes in BZDRs partially to the age-related changes in pharmacodynamics and pharmacokinetics, which increase the central nervous system (CNS) effects of these drugs in older persons.

Saarelainen added, "Unfortunately, we couldn't study the question of the mechanism of death in these individuals. But based on previous studies, the deaths may be due to other outcomes, such as fractures and pneumonia, which are fatal in this population."

Nonpharmacologic Interventions

Commenting on the study for Medscape Medical News, Charles F. Reynolds III, MD, professor of geriatric psychiatry, University of Pittsburgh, and director of the Aging Institute, University of Pittsburgh Medical Center, Pennsylvania, said that the article "does a nice job of documenting the association between benzodiazepine exposure and mortality risk in the context of AD."

However, he expressed concern that "interpretation of data was not altogether straightforward because behavioral and psychological disturbances in the context of dementia themselves confer greater risk of early mortality.

"That said, the association is certainly plausible, and I can invoke several potential mediating factors, such as falls, disturbances in respiration during sleep, and cardiac arrhythmias," Dr Reynolds added.

Although the association is "real, the article does not completely convince me that we are not dealing with confounding by association but with plausible mediating variables that can explain the association, which leads to the question of whether we can associate true causal effect between benzodiazepines and mortality in AD," said Dr Reynolds, who was not involved in the study.

Regardless, there remain several important take-home messages, he added.

"Most geriatric psychiatrists try to avoid use of CNS-depressing drugs, including benzodiazepines, in patients with AD because they worsen cognitive function, increase fall risk, which can be fatal, and can worsen sleep-disordered breathing."

He suggested increased use of behavioral and nonpharmacologic approaches as the "first step in the treatment algorithm" and that other classes of medications, such as antipsychotics ― particularly quetiapine ― and antidepressants ― particularly citalopram and escitalopram ― with "a greater margin of safety" be used if these measures are insufficient.

Saarelainen also emphasized the importance using nonpharmacologic options as first-line approaches, adding that it is important to "educate caregivers to recognize the symptoms of behavioral disturbances and what might be causing these symptoms, and then address those causes."

The study received no funding. Laura Saarelainen has received a personal research grant from the University Pharmacy outside the submitted work. Other study authors have received funding from multiple sources. Dr Reynolds has received grant/research support from the National Institute of Health, the National Institute of Mental Health, the National Heart Lung and Blood Institute, the Center for Medicare & Medicaid Services, the Patient Centered Outcomes Research Institute, the John A. Hartford Foundation, the American Foundation for Suicide Prevention, the Commonwealth of Pennsylvania, the Clinical and Translational Science Institute, the National Palliative Care Research Center, and the American Association for Geriatric Psychiatry. Bristol Meyers Squibb, Forrest Labs, Lily, and Pfizer provide pharmaceutical supplies for NIH-sponsored work.

Int J Geriatr Psychiatry. Published online November 15, 2017. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: