Hypnotic medications in older adults : a call for caution

Introduction The elderly often suffer from insomnia, which can influence quality of life. Some suffer from insomnia because of ongoing medical issues; however, many suffer from primary insomnia. A few elderly adults seek treatment for insomnia by using hypnotic medications that may have more adverse effects than benefits. Several of these adverse events, like falls with subsequent injury or fracture, can be life changing. In this article, we review the various modes of insomnia treatment with a literature search on both pharmacologic and nonpharmacologic treatments. Conclusion Cognitive behavioural therapy may be an effective long-term treatment for insomnia in the elderly and is a nonpharmacologic approach. Pharmacologic treatment may be effective and efficacious in the short term. Many of the commonly used hypnotics have not been studied over a longer term. Providers should review the potential risks and benefits before initiating any type of treatment plan for insomnia in older adults.


Introduction
Insomnia, prevalent in 30% of the elderly population, is a frustrating problem that may influence quality of life 1 .In a European study of 14,915 people with insomnia (ages Hypnotic medications in older adults: a call for caution S Datta 1 , PY Takahashi 1,2,3 * 15-100 years), 18.5% were older than 65.Fifteen percent of patients had insomnia lasting 1-10 years, identified as chronic insomnia.In patients between 80 and 89 years, 23% of men and 41% of women experience insomnia 2 .People often seek medical care for their insomnia because it hinders their quality of life by increasing daytime fatigue, psychological stress and physical discomfort 1,2 .The goal of this article is to review various modalities of treatment of insomnia in the elderly.

Sleeping patterns and common causes of insomnia in the elderly
Many older adults may have lighter, less efficient and less restful sleep.Older adults often experience multiple awakenings and take more naps compared with younger adults 3 .They also report excessive daytime somnolence.The elderly endure insomnia later in life with other comorbid health conditions that may develop into secondary insomnia.Comorbid illnesses that lead to secondary insomnia may perpetuate insomnia.Providers should treat secondary insomnia by treating the comorbid illness.For example, providers could prescribe pain relievers for arthritis pain, which may result in better sleep.Chronic conditions causing insomnia are listed in Table 1.Medications can also cause insomnia, and they are listed in Table 2. Treating underlying problems and changing medications may be an effective method of reducing insomnia.

Medications
Beta-blockers, alcohol and caffeine inhibit melatonin production.Thus, people worry about not sleeping, which further disrupts their sleep 5 .Patients may require some cognitive intervention to help break these cycles.Older elders may experience emotional stress that leads to excessive worrying, which starts negative self-talk.The negative self-talk perpetuates dysfunctional beliefs about sleep that cause emotional stress, which prolongs the insomnia cycle.Cognitive training helps people cope with the negativity emanating from thoughts about sleep.By reducing negative self-talk and stress, patients can break the cycle and improve sleep.Behavioural aspects of CBT include defined steps that patients can take to improve sleep.Patients can improve sleep hygiene through stimulus control, sleep restriction and relaxation as components of behavioural therapy.These recommendations are listed in Table 3. include excessive sedation, motor in coordination, cognitive impairment and anterograde amnesia, falls due to these effects and death 14,15 .However, some studies have indicated that benzodiazepines may not be related to hip fractures, for example, no dramatic declines in the rate of hip fractures among people older than 55 years of age were observed in association with decreased benzodiazepine use 16 .There were no statistically significant differences between triazolam and temazepam for risk of hip fracture among elderly residents; risk ratio was 0.92 (95% CI, 0.72, 1.17) in a retrospective cohort study.This study compared these two shortacting benzodiazepines (triazolam and temazepam) with other sedative hypnotics and benzodiazepines.However, this study was limited by absence of a non-user insomniac control and short period of follow-up of 30 days 17 .In patients over 65 years of age, 245 cases were matched to 817 controls; hip fracture was, however, associated with the use of two or more benzodiazepines, but, only lorazepam was significantly associated with an increased risk of hip fracture (risk ratio of 1.8, 95% CI,1.1, 3.1) 18 .On the contrary, a meta-analysis of 24 RCTs in the USA with 2417 participants with a mean age of 60 showed an increased risk of adverse events with the use of benzodiazepines.Cognitive adverse events were 4.78 times more (95% CI, 1.47, 15.47); adverse psychomotor events were 2.61 times more (95% CI, 1.12, 6.09) in patients on hypnotics than placebo 14 .Total sleep time increased by a mean of 25.2 minutes (P < 0.001).
The number needed to treat insomnia with hypnotics was 13.Unfortunately, patients in hypnotic group suffered more adverse events with a number needed to hospitalise after hypnotic use of 6 14 .Patients and providers fear falls and increased mortality as the most serious adverse side effects of hypnotic agents, especially benzodiazepines.In a case-control study of Both short-and long-term benefits of CBT have been proven by several randomised control trials (RCTs) 6,7 .An RCT studied self-help CBT in 193 patients between ages 55 and 87 with long-term conditions and chronic insomnia symptoms.Significant increases in total sleep time were reported for the intervention group at post-treatment (adjusted mean increase in sleep time of 33.1 minutes, P < 0.001), 3-month followup (adjusted mean increase in sleep time of 39.5 minutes, P < 0.001) and 6-month follow-up (adjusted mean difference = 41.4 minutes, P < 0.001) 7 .In a large meta-analyses of patients with a mean age of 55 years, total sleep time improved by 17% (95% CI, 0.01-0.33)and sleep efficiency (SE) improved by 55% (95% CI, 0.28, 0.75) 6 .
Although CBT is an important modality, providers and patients are also looking at phototherapy to improve sleep patterns.For effective phototherapy, one requires the natural intensity of light on a sunny day to take a photo without a flash.An older patient would require 2500-3000 lux for 20-30 minutes daily for 3-4 weeks per expert consensus for an effective phototherapy session 8 .In an RCT of 115 patients, light participants were awake 39 fewer minutes (95% CI, −76.0 to −1.9 minutes; P = 0.04; effect size = 0.53) with 30 minutes of 2500 lux of light therapy daily 9 .Light therapy using a light box is generally a safe option with infrequent side effects including headache, nausea and eye irritation.
Exercise is another potential nonpharmacologic option to help with sleep.For 3 years, 10,211 older adults were studied for frequencies of physical activity and insomnia.Physical activity for five or more days per week reduced the incidence of insomnia, especially for difficulty maintaining sleep 10 .Another study of 29 elders with a mean age of 88.3 years observed that nighttime sleep percentage increased from 51.7% at baseline to 62.5% after 14 weeks of exercise for 20-30 minutes, 4-5 days of the week 11 .
Other options available are acupuncture and transcutaneous electrical nerve stimulation.Three times per week for 3 weeks of electroacupuncture of primary insomnia was studied randomised, single-blind, placebo-controlled, the proportions of subjects having less than 30 minutes of wake after sleep onset and an SE of at least 85% at the post-treatment visit were significantly higher in the electroacupuncture group (73% in placebo and 81% in intervention group) 12 .These findings were based only on subjective evidence and not supported by a polysomnogram.More randomised control trails are needed to support the efficacy of transcutaneous electrical stimulation when compared with the sedative-hypnotic medications.

Pharmacologic treatment
The pharmacologic treatment of insomnia in older adults poses a greater dilemma for clinicians, as patients desire a rapid solution for insomnia.Older adults are more than twice as likely to be prescribed a sedative/hypnotic medication for insomnia, as are younger adults 13 Clinicians often desire to reduce the discomfort of insomnia; however, patients often face serious side effects of sedative/hypnotic medications (hereafter called hypnotic medications).In elderly, altered pharmacokinetics of medications leads to delayed metabolism of medications.Delayed metabolism can lead to medication or metabolite accumulation, which impairs cognitive skills, psychomotor skills and balance.Long-term use of hypnotic medication for insomnia lacks evidence of efficacy.Providers and patients should strive to reduce long-term use of hypnotic medications because of the inherent risks.
Benzodiazepines are popular among the elderly for insomnia.The side effects of benzodiazepines with a 10-year follow-up of 3777 patients with a mean age of 79 years, the odds ratio of falls in older adults using benzodiazepine compared with non-use was 2.2 (95% CI, 1.4, 3.4) in the population >80 years.In younger patients under 80, the odds ratio was 1.33 (95% CI, 0.9, 1.9), which was not significant 19 .In a 15-year follow-up of 3777 community-dwelling elders with a mean age of 78.2 years, new use of benzodiazepines was associated with an increased risk of dementia (multivariable-adjusted hazard ratio 1.60; 95% CI, 1.08, 2.38) 20 .
GABA receptor agonists, also popularly known as Z-drugs, are some of the most widely used hypnotic medications.In clinical practice, 6 months of zolpidem treatment is shown to be efficacious and safe 21,22 .In a doubleblind, placebo-controlled randomised PSG trial, the first of its kind, evaluating 8 months of nightly zolpidem treatment in primary chronic insomniacs, Zolpidem significantly increased SE increased from 79% to 85% at 1 month and maintained at 85% at eighth month, when assessed at months 1 and 8 relative to baseline and placebo 22 , but the side effect profile may limit the use of these in the elderly.The side effects of GABA receptor agonists are similar to benzodiazepines.The side effects include cognitive impairment, somnambulism, nightmares and sensory disturbances 23,24 .In a study of 41,978 inpatients with a mean age of 56.8 years, zolpidem use compared with non-use was independently associated with an increased risk of inpatient falls with an adjusted odds ratio of 4.37 (95% CI, 3.34, 5.76; P < 0.001) 25 .Fractures, as an outcome of falls, are another devastating issue due to their impact on a person's independence and the financial burden on the health care system.In a study of 43,000 patients, the rates of nonvertebral fractures pre-and post-treatment with zolpidem were similar in patients younger than 65 years of age (5-10 per 1000 person years); however, the rate of fractures increased from 10 per 1000 years in patients aged 65-69 years to 25 per 1000 years inpatients aged > 75 years (post-treatment with zolpidem, P = 0.04) 26 .In addition, mortality may increase with hypnotic medication use.In a study of 10,529 patients with a mean age of 54 years, investigators found that users of hypnotic prescriptions have a greater than threefold increased hazard of death with hypnotic medications compared with those without, even when prescribed <18 hypnotic pills per year.The hypnotic prescriptions in this study included zolpidem, temazepam, eszopiclone and zaleplon, other benzodiazepines, barbiturates and sedative antihistamines 27 .
Despite their adverse effects, hypnotics continue to be prescribed.In 2011, 39 million prescriptions for zolpidem products were prescribed 28 .A meta-analysis of 137 studies with 1276 individuals compared the efficacy of benzodiazepine and GABA receptor agonists.The authors reported that total sleep time improved by 49.2 minutes (95% CI, 36, 62.5) with triazolam and 32 minutes (95% CI, 21.7 to 42.3) with zolpidem.After 1-3 nights of medication withdrawal, total sleep time decreased by 17.7 minutes (95% CI, -42.9, -7.6) with triazolam and 2.4 minutes (95% CI,11.4,6.6) with zolpidem.Hence, effects in efficacy or rebound insomnia of both the benzodiazepines and non-benzodiazepines were comparable 29 .Rebound insomnia can be decreased by gradual dose reduction over months or weeks 30 .In January 2013, FDA officials recommended changing the dosing of zolpidem due to its delayed metabolism in women.Women should take half the dose of zolpidem compared with men.Zolpidem tartrate, a medication for middle-ofthe-night awakening was studied in 295 patients with primary insomnia and difficulty returning to sleep after awakenings.It decreased latency to sleep onset by 30 minutes as compared with placebo (P < 0.0001) 31 .A steep surge in emergency department visits was seen due to adverse reactions to the sleep medication zolpidem (220% from 2005 to 2010, from 6111 visits to 19,487 visits).Fifty percent of these were due to zolpidem use with other drugs, and 37% of all emergency department visits related to zolpidem involved additional use of drugs that depressed the central nervous system, such as antianxiety medications such as benzodiazepines, narcotic pain relievers and alcohol 32 .Hence it is important that prior to prescribing these medications, the drug interactions be considered, and close monitoring and follow-up be arranged.
Another commonly used hypnotic agent is trazodone.Trazodone, a weak 5HT inhibitor and alpha blocker, is an antidepressant with sedative properties.It is not FDA approved for insomnia but is widely used off-label.Trazodone 50 mg was administered to participants 30 minutes before bedtime for 7 days in a 3 week, in a randomised, double-blind, placebocontrolled trial, trazodone decreased night-time awakenings (from 16 to 12).In addition, simulated driving and equilibrium was unaffected by trazodone.In general, trazodone is well tolerated, and the adverse events are transient and mild.Trazodone is commonly prescribed in 25-50 mg doses for insomnia.In the largest randomised, double-blind, parallelgroup trial, 306 patients used trazodone or zolpidem for 14 days.Both agents reduced sleep latency by 15 minutes, but this reduction in sleep latency for trazodone lasted 1 week while zolpidem lasted 2 weeks 33 .Side effects of trazodone include drowsiness, dizziness, constipation, blurry vision, confusion, hypotension, priapism (urologic emergency) and serotonin syndrome with selective serotonin reuptake inhibitors (SSRIs) and selective norepinephrine reuptake inhibitors (SNRIs) 34 .Patients who are at high risk for developing priapism include those with sickle cell anaemia or sickle trait, leukaemia, autonomic nervous system dysfunction or hypercoagulable states 35,36 .The effects of trazodone on cognition have not been substantially proven; hence trazodone remains a viable option for the elderly if the risks and benefits are weighed.
Other hypnotic agents commonly prescribed include tricyclic antidepressants.Doxepin is an FDA-approved tricyclic for insomnia.It has been shown to improve sleep by increasing SE and total sleep time and decreasing sleep latency 37 .In a clinical trial with doxepin, 240 patients greater than 70 years showed that low-dose doxepin (1 or 3 mg) improved sleep latency by 12 minutes, increased total sleep time by 25 to 51 minutes and increased SE by 6%-10% 37 .The safety profile of low-dose doxepin (<6 mg) is comparable with placebo.The anticholinergic effects of doxepin appear with a dosing of 10 mg and above, its antidepressant dose 37 .Mirtazapine (a tetracyclic antidepressant) has sedating properties.It can also cause weight gain and improve appetite in addition to causing sedation effects 38 .
Even though it is given at bedtime when used as an antidepressant, its primary insomnia is discouraged.However, in depressed patients, mirtazapine has been reported to improve sleep architecture and SE 39,40 .Hence it could be used to treat insomnia in depressed patients.
First-generation antihistamines are commonly used hypnotic agents in over-the-counter medications.The American Academy of Sleep Medicine does not recommend antihistamines for the treatment of insomnia, as daytime drowsiness can be severe in the elderly, even when the medication is taken the prior day or night.Tolerance to the drug can develop in as little time as 3-4 days.Anticholinergic side effects in older adults needing medical attention may be dizziness, orthostatic hypotension, confusion, blurry vision, urinary retention and falls 41 .A randomised, controlled, crossover clinical study compared 14 nights of treatment with 15 mg temazepam, 50 mg diphenhydramine and placebo in elderly individuals with insomnia (mean age, 73.9 years).Improvements were seen with diphenhydramine treatment compared with placebo on the number of awakenings only, but there were no effects on total sleep time or sleep latency 41 .Diphenhydramine does have an FDA indication for insomnia.In a study of 105 patients comparing sleep quality with diphenhydramine and placebo, decreased sleep latency, decreased awakenings and increased duration of sleep were found with diphenhydramine as compared with placebo (P < 0.002) 42 .
Patients are also turning to nutraceuticals to help with sleep.Melatonin, a hormone released from the pineal gland, helps control the circadian rhythm.Melatonin is available in sustained-release formulation of 1, 2 and 3 mg and is given 2 hours prior to bedtime.In a systematic review of 16 RCTs, melatonin decreased sleep onset latency by 11.7 minutes (95% CI, -18.2, -5.2).There was no evidence of adverse effects of melatonin.Clinical trials with 12 studies of patients 65 years and older showed decreased sleep onset latency by 11.7 minutes 43 .Melatonin is relatively safe; however, some patients have reported fatigue, headache, irritability and dizziness 44 .The melatonin receptor agonist ramelteon is an FDAapproved agent but is contraindicated with liver failure patients.An RCT of 107 patients aged 18-64 years resulted in statistically significant reductions in sleep latency by 15%-20%, which is equal to the range of total sleep times of other hypnotics.There were no next-day residual effects measured by mood and feeling and immediate recall and delayed recall 45 .An RCT exam-ining the effects of ramelteon versus placebo with zolpidem studied balance in older adults after middle of the night awakening.Ramelteon did not affect balance and mobility compared with the placebo, although 10 mg of zolpidem did show statistically significant impairment.Additionally, the number of falls experienced by those taking ramelteon and the placebo were similar, whereas those who received zolpidem had a much higher frequency of falls 46,47 .

Conclusion
Insomnia in older adults is a common and challenging problem for clinicians.Every effort should be made to use nonpharmacologic methods to improve sleep hygiene compared with pharmacotherapy as CBT has been repeatedly proven to be more effective with minimal side effects.Some commonly used sleep medications do not have FDA approval for use in insomnia.In those medications with FDA approval, there are significant side effects of these medications in older adults, the primary side effect being falls with potential ensuing fractures.If these medications are used in the elderly, they should be closely monitored for interactions and side effects.Attempts to taper these medications should be done frequently.Tapering hypnotic medications over an extended time might be a wise course of action for providers.Educating patients on the side effect profile and reinforcing the behavioural techniques would help decrease the need for sleeping pills; hence promote more safety in the elderly population.