Benzodiazepine Use Linked to Dementia Risk

Megan Brooks

October 02, 2012

October 2, 2012 — Older adults who use benzodiazepines have about a 50% greater chance of developing dementia than their peers who don't use benzodiazepines, French researchers observed in a large population-based study.

This finding, added to evidence of increased risk for falls and fractures in elderly who use benzodiazepine, "should incite (health providers) to carefully assess expected benefits versus putative risks, and to limit prescriptions to a few weeks," lead author of the study and PhD student Sophie Billioti de Gage, PharmD, from the University of Bordeaux Segalen in France, told Medscape Medical News.

"In any case, uncontrolled use should be cautioned against," she said.

Dr. Sophie Billioti de Gage

The study was published online September 27 in the British Medical Journal.

"Bad Drugs for Older Adults"

Greg A. Sachs, MD, who reviewed the study for Medscape Medical News, said it is "yet another study that suggests that benzodiazepines are bad drugs for older adults."

Dr. Sachs is chief of the Division of General Internal Medicine and Geriatrics, Indiana University School of Medicine, and investigator at the IU Center for Aging Research, Regenstrief Institute, Indianapolis. He was not involved in the study.

"Many of these drugs," Dr. Sachs said, "are on 'Do Not Prescribe' lists for older adults. If used at all, they should be used for short periods of time (10 days or less) and in the lowest dose possible to achieve benefit for the target symptom." Benzodiazepines "should not be used long term for either sleep or anxiety; safer alternatives exist."

Dr. Greg A. Sachs

The analysis included 1063 men and women (mean age, 78.2 years) from the PAQUID (Personnes Agées Quid) project, a prospective, population-based study of cognitive aging and dementia involving a total of 3777 participants from France. The study started in 1987 and follow-up lasted 20 years, with clinic visits every 2 to 3 years.

All participants in the current analysis were free of dementia at the outset and did not start taking benzodiazepines until at least the third year of follow-up. Ninety-five (8.9%) reported benzodiazepine use at the 5-year visit, indicating new use between years 3 and 5. Year 5 was baseline for the analysis.

During the 15-year follow-up period, 253 (23.8%) cases of dementia were confirmed: 30 (32%) in benzodiazepine users and 223 (23.0%) in nonusers.

The 15-year incidence rate of dementia per 100 person-years was higher in benzodiazepine users than nonusers (4.8 vs 3.2).

The multivariable adjusted hazard ratio (HR) for dementia with new use of benzodiazepines was 1.60 (95% confidence interval [CI], 1.08 - 2.38). This result was unchanged when further adjusted for depressive symptoms (HR, 1.62; 95% CI, 1.08 - 2.43).

The result "remained robust" in a secondary pooled analysis of patients who initiated a benzodiazepine between follow-up visits 8 and 15 (HR, 1.46; 95% CI, 1.10 - 1.94). This added a total of 116 additional new users during follow-up to the 95 new users at year 5.

Similarly, in a nested-case control study (467 case-patients with dementia and 1810 controls), the adjusted odds ratio (OR) with ever use vs never use was 1.55 (95% CI, 1.24 - 1.95). The results were similar in past users (OR, 1.56; 95% CI, 1.23 - 1.98) and recent users (OR, 1.48; 95% CI, 0.83 - 2.63).

The PAQUID investigators say their findings are consistent with 3 recent case-control studies that found an increased risk for dementia in benzodiazepine users.

Two of the studies from Taiwan (Wu et al, 2011, Wu et al, 2009) used health insurance data and showed an increased risk for dementia in long-term users ( > 6 months; adjusted OR, 1.24; 95% CI, 1.01 - 1.53) and current users (adjusted OR, 2.71; 95% CI, 2.46 - 2.99).

The third study, a nested case-control study among French people, found an increased risk for dementia in former users (adjusted OR, 2.3; 95% CI, 1.2 - 4.5).

Other studies, however, have not found an increased risk for dementia among elderly people using benzodiazepines.

Dr. Billioti de Gage told Medscape Medical News that "contrary to most of the previous study on the topic, our study is based on a long period of follow-up (up to 15 years) and was carried out in a large representative cohort of elderly participants. This allows to take into account the somewhat long prodromal period of dementia and to generalize the conclusions."

"Positive Distinguishing Factors"

Dr. Sachs said there are several "positive distinguishing factors about this study." It was large; it followed patients over many years with little dropout; it was prospective and longitudinal rather than cross-sectional; the diagnosis of dementia involved both neuropsychological testing and examination by a neurologist; and the researches had excellent information about drug use from patients, he explained.

The exclusion of people who were already receiving benzodiazepines at time of study entry and for a period of a few years "run in" is another key strength, he said.

"This is important regarding the notion of 'reverse causation' — that people could end up taking benzodiazepines because of symptoms of depression or anxiety that are early symptoms relating to a developing dementia. Without doing that, it could bias the study toward people with dementia already brewing getting benzodiazepines at a higher rate, instead of the notion that it is contributing to dementia development," Dr. Sachs said.

The analyses were "carefully done" and the findings were "explored and confirmed using more than one approach. The PAQUID study is one of the higher quality cohort studies examining dementia," he added.

Dr. Billioti de Gage said, "patients should be told about the potential adverse effects of these drugs, including long-term risk when initiating benzodiazepines and about the necessity of gradual discontinuation when stopping the treatment."

She and her colleagues say further study is needed to determine whether long-term use of benzodiazepines in people younger than age 65 years is also associated with an increased risk for dementia and uncover possible correlations between dosage or cumulative length of exposure and dementia.

Dr. Sachs agrees. The analysis "cannot tell us anything about long versus short acting meds, specific meds, dose, or duration of therapy," he told Medscape Medical News. Also, the small numbers of people on benzodiazepines in the study is a limitation, he added.

Another limitation, he said, is that the analysis is primarily focused on "ever use" of benzodiazepines, "and that means we do not have information here on whether stopping the meds would help prevent dementia. In fact, because of the 'ever use' approach to analysis, I'd be concerned that someone misinterpret this and think 'why bother stopping' once someone has been exposed."

Dr. Sachs also noted that "far greater numbers of people who developed dementia had not been exposed to benzodiazepines than those who had; so while it increases relative risk, how much it contributes to development of dementia should not be overplayed; this still was an observational study, so assigning causation is hard to do no matter how well the study is done."

This research was conducted by the INSERM U657 research team co-funded by INSERM (Institut National de la Santé et de la Recherche Médicale) and Université Bordeaux Segalen. Additional support was provided by a 2010 grant from IRESP (Institut de Recherche en Santé Publique) acting on behalf of the French Ministry of Health (Direction Générale de la Santé, Direction de la Recherche, des Études, de l'Évaluation et des Statistiques); by a 2011 grant from the French Ministry of Health (Direction Générale de la Santé); and by Caisse Nationale des Travailleurs Salariés, Régime Social des Indépendants, Caisse Nationale de Solidarité pour l'Autonomie, and Institut National de Prévention et d'Education pour la Santé. SBdG is a part-time researcher in the INSERM 657 Unit, and her salary is paid by IRESP.

BMJ. Published online September 27, 2012. Abstract