Benzodiazepines in elderly patients

For a long time, benzodiazepines have provided an effective treatment of anxiety disorders and sleep disturbances.¹ They were also found to be helpful in anxious depression, bipolar disorder, catatonia, and in a number of medical conditions, such as epilepsy and alcohol withdrawal.¹ In the 90s, introduction of second-generation antidepressants provided more expensive modalities of addressing anxiety disorders. Substituting benzodiazepines with selective serotonin reuptake inhibitors (SSRI) and serotonin-noradrenaline reuptake inhibitors (SNRI) clearly appeared the commercial way to go. Such road would have been difficult to follow when new medications had to be compared with a gold standard, because direct comparisons clearly indicated higher efficacy and tolerability of benzodiazepines over antidepressants.² However, when such superiority was no longer required by regulatory agencies, comparisons were performed by meta-analytic methods that are liable to manipulation instead of head-to-head.³ The other complementary strategy was to magnify side effects of benzodiazepines. As the negative aspects are actually very limited,¹ the potential for dependency, toxicity, and abuse of benzodiazepines had to be dramatically pictured.⁴ Not surprisingly, a decade ago benzodiazepines were viewed in guidelines and regulatory bodies as having very limited benefits and considerable risks, suitable only for short-term treatment.^{1, 4} Recently, however, a reconsideration of their role and function has occurred. First, studies in pharmacoepidemiology indicate that benzodiazepines are seldom misused or lead to misuse of other substances in patients without substance use disorders who are prescribed these medications for appropriate indications.^{1, 4} Second, withdrawal reactions may occur after discontinuation of both benzodiazepines and antidepressants, but for the latter such events appear to be more troublesome and persistent.⁵ Third, the ghost of dementia evoked in observational studies did not manifest its presence when carefully conducted methodological investigations examined the association.^{6, 7} The paper by Rozing et al.⁸ adds a fourth important pillar, challenging the associations between benzodiazepines and the risk of falls and fractures.

Correlational methods in epidemiology entail the risk of yielding spurious results when studying highly heterogeneous constructs and populations. Confounding by indication appears to be particularly challenging in assessing medication risk and falls. It occurs when a person takes a medication (e.g., benzodiazepine) for a disorder (e.g., insomnia), which itself can increase the risk of falling (by waking up at night and walking in the house). Rozing et al.⁸ used a self-controlled case series design where different time periods are considered: baseline unexposed to hypnotic-sedative drugs (24–12 months before initiation); early pretreatment (12–3 months before beginning); pretreatment (3 months immediately preceding the purchase of the medication); treatment (the first 3 months); early posttreatment (3–12 months after treatment initiation); late posttreatment (12–24 months after treatment initiation). Such design may minimize, even though not completely eliminate, confounding by indication. Nearly 700,000 adults with a purchase of benzodiazepines, Z-drugs (medications acting on benzodiazepine receptors that do not have a benzodiazepine structure, i.e., zolpidem and zopiclone) and melatonin in the Danish National Patient Registry between 2000 and 2018 were included. A fall accident occurred in 8.7% of the study population and a fracture in 5.2%.⁸ Their occurrence was likely to be underestimated, because it included only hospital diagnoses. As expected, older age (equal or above 70 years) was found to be related to such occurrences.⁸ For older people, the risk of falls was highest during the pretreatment period, suggesting that their increased risk may be due to factors other than benzodiazepines, Z-drugs, and melatonin, particularly untreated insomnia. The initiation of treatment gradually diminished the risk.

Can we conclude that hypnotic-sedative medications are devoid of risks associated with falls and fractures and we should not worry about the issue? Not at all: common side effects of benzodiazepines and Z-drugs, such as sedation, slowed reaction time, impaired balance and gait, may indeed facilitate falls, even though no more than other medications. Rozing et al.,⁸ in line with the available literature,¹ underscore the benefits of addressing anxiety and sleep disturbances in the elderly, particularly in the setting of medical disease. An updated consideration of the literature indicates that benzodiazepines are first-line medications for addressing anxiety and sleep disturbances, even though non-pharmacological strategies such as cognitive behavior therapy and lifestyle modifications should have a primary role.^{1, 2, 4}

Falls and fractures are a major source of morbidity and mortality in older patients, triggering a cascade of adverse events, such as pulmonary infections.⁹ Preventing or minimizing the risk of falling deserves clinical attention in managing anxiety and insomnia. Such consideration requires a personalized approach that integrates the mere use of diagnostic criteria leading to automatic prescriptions.^{3, 10} I will use the common and vexing problem of insomnia in elderly patients to illustrate how a personalized approach can be implemented, using examples from my extensive clinical experience.

Not all patients over 70 years of age present the same risk of falling: strength, gait, and balance impairments, in addition to the occurrence of previous falls, are strong predictors of falling.¹¹ In the real world of multimorbidity and polypharmacy, which is quite different from the one depicted by clinical guidelines and evidence-based medicine,³ we need to evaluate the overall medical and functional condition of the patient (the geriatric syndrome of frailty reflects this global evaluation), as well as the cumulative effect of the medications that are taken.¹¹ Whenever appropriate, deprescribing should be considered,¹¹ such as in the case of antidepressant drugs that may increase the risk of falling because of specific side effects (e.g., orthostatic hypotension).^{10, 12} Antidepressant medications, however, are required when early morning awakening associated with a major depressive episode occurs.¹⁰

A second aspect that has to be taken into consideration is the fact that not all benzodiazepines are the same and treatment choices should be carefully pondered. Elderly patients are generally more responsive to benzodiazepine effects as a result of increased adipose/lean ratio and decreased plasma proteins, resulting in slowed metabolism, decreased clearance, and longer elimination half-life.¹ As a result, caution is needed when using benzodiazepines with longer half-lives because of the risk of accumulation.¹ Moreover, half-life characteristics need to be integrated with relative lipid solubility and binding affinity.¹³ Drugs like triazolam and alprazolam, which have high lipid solubility, are associated with rapid effects, higher dependence liability, cognitive impairment, and anterograde amnestic effects.¹³ On the contrary, benzodiazepines with low affinity for the benzodiazepine receptor and lipid solubility, such as clonazepam, entail slower effects, less dependence liability, and amnestic potential.¹³ In a 12-year national-registry study performed in Luxembourg, about four out of five people who received a benzodiazepine were short-term or intermittent users.¹⁴ Continuous use, not necessarily associated with dose escalation, occurred in the remaining cases. Alprazolam and triazolam were related to continuous and high-dose use, whereas clonazepam and clobazam were not.¹⁴ The findings of this investigation support the importance of a specific benzodiazepine selection based on the likelihood of dependence potential and on the specific targets of the prescription.¹³ Dosage may be particularly important as to the risk of falling and fractures. High doses of hypnotics at bedtime are often used, but this practice could also involve an increased risk of falling at night. Even though insomnia may be the presenting symptom, the level of anxiety during the day needs to be evaluated. As a patient once shared with me “it is the day that dictates the night; in certain days when my tension is so high probably only general anesthesia at bedtime would work for me.” I thus tend to avoid the use of hypnotics altogether and favor splitting a benzodiazepine during the day. Very often I have been able to successfully address insomnia that was refractory to hypnotics by switching the patient to very low-dose clonazepam (0.25 mg in two divided doses), with the benefit of avoiding a medication bulk at night.

A final source of consideration is that the use of benzodiazepines in insomnia should be placed within the context of lifestyle medicine.¹⁵ Suggestions geared to a better sleep need to be individualized and address unhealthy practices during daytime (e.g., insufficient physical activity, prolonged and/or frequent napping, getting home very late from work), before going to sleep (e.g., caffeine consumption, falling asleep in front of the TV), and during the night (e.g., how to get up from bed, how to go safely to the bathroom). At times, lifestyle suggestions may help avoiding the use of medications; more frequently they may increase their effectiveness and facilitate their tapering and discontinuation. If I write a prescription, I always add a second prescription with lifestyle suggestions, emphasizing that it is at least as important as the first.¹⁰

The optimal pharmacological treatment of insomnia should be short also in elderly patients. However, this is difficult in many cases. Fortunately, benzodiazepines do not appear to lose effectiveness with time and are unlikely to require increasing the dose.¹ You may find, for instance, patients who have been getting a good night sleep by taking 1 mg of lorazepam over many years. The problem is when, for whatever reason, they stop taking it. Very severe withdrawal syndromes,⁵ including delirium, might ensue.

In conclusion, the findings of the important epidemiological study by Rozing et al.⁸ add to other evidence,¹ suggesting that benzodiazepines do not increase the risk of falling and fractures more than other medications. However, preventing such occurrences should be a primary source of concern for physicians dealing with anxiety and insomnia in elderly patients.

The author declares no financial conflict of interest.