*Harvard Mental Health Letter* includes an
article: “Insomnia in later life: Overcoming obstacles to a good night’s
Here’s the article:
The older you are, the more likely you are to have a sleep disorder.
According to a National Institute on Aging study, more than 50% of
people over age 65 report regular sleep problems that trouble them at
night or interfere with daytime activities. They can’t fall asleep when
they want to, they wake up repeatedly, they wake up too early, their
sleep is not refreshing, or they feel drowsy or groggy all day.
Fortunately, our understanding of sleep itself and the sleep disorders
of late life have improved greatly, and help is more easily available
Sleep becomes shallower and briefer with age. Its deepest stages
practically disappear by age 50 in many people. As we grow older, we not
only sleep more lightly but wake up more often, have more brief periods
of wakefulness, and spend more time lying awake (experts use the term
“poor sleep efficiency”). We may fall asleep and wake up earlier than is
desirable, and make up for sleepless nights, if we can, by daytime napping.
One result is falling asleep at the wheel, in the metaphorical sense
(loss of alertness, concentration, and memory) and the literal sense
(drowsiness may account for 100,000 automobile accidents each year, many
of them involving older drivers). Insomnia may be a mortal danger in
other ways as well. According to a study, people who sleep less than
five hours a night have twice the average rate of heart attacks. Another
study found that older adults who lay awake in bed for a half-hour or
more on most nights had a high death rate even if their overall health
Insomnia has many causes. Medical conditions that can lead to sleep loss
include overactive thyroid, diabetes, congestive heart failure, high
blood pressure, asthma, emphysema, arthritis pain, chronic heartburn,
and urinary difficulties. Some potentially insomnia-provoking
medications and drugs are alcohol, caffeine, stimulants, steroids,
diuretics, cold and allergy medications, antidepressants and anti-
Insomnia and daytime drowsiness are also features of many sleep
disorders that occur in middle age and later. The most common is sleep
apnea, which occurs when a person’s breathing stops many times during
the night, often for a minute or more at a time, because the airway is
blocked by sagging tissue or the brain does not reliably cue breathing
muscles in the diaphragm. People with sleep apnea wake up momentarily
every time their breathing stops. Their sleep is not refreshing, and
they may suffer morning headaches as well as daytime drowsiness. Their
hearts bear the strain of working to supply oxygenated blood to the body
during long periods when they are not breathing. The risk is greatest
for overweight men, especially if they drink alcohol in the evening.
Circadian rhythm changes are another common feature of later life. The
internal clock that controls the 24-hour cycle of sleep and waking
drifts forward (i.e., earlier) in older people, so they are more likely
to fall asleep early in the evening and wake up in the middle of the
night. If they try to stay up later, they may be sleepy all day. The
condition is called advanced sleep phase syndrome.
Insomnia and psychiatric disorders
Psychiatric conditions are a major cause of insomnia in later life. The
Epidemiologic Catchment Area (ECA) survey of nearly a thousand Americans
found that about half of those with insomnia had a psychiatric disorder,
most commonly an anxiety disorder, depression, or alcohol abuse or dependence.
About two-thirds of depressed patients sleep too little. Their most
common sleep symptom is waking up too early, but they also have less
deep sleep, and often their rapid eye movement (vivid dreaming) sleep
begins abnormally early in the night — a phenomenon called short REM
latency. Studies also show that depressed people often think they get
even less sleep than they do. Meanwhile, a minority who suffer from
winter depression (seasonal affective disorder) or the depressed phase
of bipolar disorder sleep too much rather than too little.
This connection runs both ways. A person may be anxious or depressed
because he is lying awake in bed as well as wakeful because he is
anxious or depressed. In the ECA survey, 17% of people who had insomnia
for a year developed a psychiatric disorder in the following year. In
another study, young people who said they had been sleepless at some
time for at least two weeks were four times more likely than average to
develop major depression in the following three years. Sleep loss also
interferes with recovery from medical and psychiatric conditions. A
person who is no longer depressed but still insomniac has five times the
average risk of relapsing.
Depression, anxiety, and insomnia may have common causes in overactivity
of brain arousal mechanisms and malfunctioning of the stress hormone
system governed by the hypothalamic pituitary adrenal (HPA) axis. So
insomnia might be an early indicator of depression and anxiety
disorders, a symptom of those disorders, or an independent effect of
similar underlying mechanisms.
Illness, emotional stress, or a change in the environment often cause
temporary sleeplessness. In about 5% of these cases, the problem becomes
chronic. This kind of insomnia, not linked to any other medical or
psychiatric disorder and therefore labeled “primary,” may be the most
common sleep disturbance of later life. It’s also called learned or
conditioned insomnia because it results from an association the mind
makes between wakefulness and being in bed — an association strengthened
and prolonged by anxiety about sleeplessness itself.
In a sleep laboratory, patients are wired to sensors that measure
various brain and body functions, including brain electrical activity,
muscle tension, airflow, heart rate, and blood oxygen. Occasionally, a
person who only thinks she sleeps poorly may be reassured by laboratory
results, but everyday (or every night) insomnia almost never requires
this technology. It’s useful mainly when there are signs of sleep apnea
or another disorder that has a special treatment.
Drugs: The easy way? By this time, almost everyone has seen
advertisements for prescription sleeping pills, and more people are
using them than ever. Prescriptions for sedative-hypnotic drugs (the
medical term) doubled between 2000 and 2004, mainly because of new
medications that are supposed to be safer and more effective than the
old. But there is still much evidence that sleeping pills are often
unreliable and can be risky, especially in the long run and for older people.
Benzodiazepines. These drugs, also used to treat anxiety and seizures,
once accounted for most sleep medicine prescribed by physicians. They
boost the activity of the inhibitory neurotransmitter gamma-aminobutyric
acid (GABA), and include triazolam (Halcion), temazepam (Restoril),
lorazepam (Ativan), and clonazepam (Klonopin). Today, they are used less
in the treatment of insomnia, replaced by benzodiazepine receptor
BRAs. Although they act on the brain in a similar way, these drugs —
zolpidem (Ambien), zaleplon (Sonata), and eszopiclone (Lunesta) — are
believed by some to be more selective than the benzodiazepines, with
fewer side effects. Compared to benzodiazepines, they may create less
risk of rebound insomnia, withdrawal reactions, and addiction.
Eszopiclone is the only sleeping pill approved by the FDA to be used for
as long as six months. Because it has a longer period of action than the
others, it may be better for staying asleep (as opposed to falling
asleep), which can be particularly difficult for older people and those
The risks of benzodiazepines and BRAs, especially daytime carryover
effects, are greater in older people because they have less resilient
bodies, and break down and eliminate drugs more slowly.
Ramelteon (Rozerem). This drug is a new kind of sleeping pill. It
differs from benzodiazepines and BRAs in its mechanism of action,
working at receptors for the hormone melatonin. So far ramelteon has
shown little evidence of causing rebound insomnia or a withdrawal
reaction. But we won’t know much about its long-term effects until it
has been used for a longer time by more patients.
Antidepressants. Antidepressant drugs with sedative properties, such as
amitriptyline (Elavil) and trazodone (Desyrel), are often used in place
of BRAs because they are less addictive. But they take several weeks to
work, and there are few controlled studies showing that they are
actually effective in the treatment of insomnia. For people with both
depression and insomnia, adding a sedative may help while they wait for
the antidepressant to take effect. One study found that a combination of
eszopiclone with fluoxetine (Prozac) relieved symptoms of both
depression and insomnia faster and more effectively than either drug
Over-the-counter medicines. An antihistamine (usually diphenhydramine)
is the main active ingredient in sleep remedies sold under names like
Benadryl, Sominex, Nytol, and Tylenol PM; sometimes a pain medicine is
added. The effectiveness of these drugs doesn’t match their popularity.
They don’t work well or for long as a treatment for insomnia, and older
people are especially susceptible to their undesirable side effects,
including daytime grogginess and blurred vision. Most experts do not
Alternative remedies. Various herbal medicines are sold as “natural”
sleep aids; the most common are valerian, kava, and chamomile. There is
little solid evidence about these herbs, and what there is suggests that
they are not particularly effective. They contain a variety of chemicals
with uncertain combined effects, and most of them have not been
carefully tested on older people.
The hormone melatonin is marketed as a dietary supplement for the
treatment of insomnia, but studies suggest that it is not effective. It
may have some promise specifically for older people with low levels of
melatonin or advanced sleep-phase syndrome.
The FDA does not regulate herbal medicines and dietary supplements, so
quality and purity may be questionable, and side effects have not been
carefully recorded or described.
Sleeping pills: Illusions?
Even at best, sleeping pills are only modestly effective. A 2005 review
based on 24 controlled trials including more than 2,000 patients found
that benzodiazepines and BRAs sometimes did more harm than good.
Compared with placebos, the drugs lengthened sleeping time by an average
of a half-hour and reduced the number of times a person woke up at night
by an average of 0.63. But the researchers estimate that 13 people had
to be treated for one to show significantly improved sleep quality, and
one in six would have some harmful effect.
So this apparently easy way to help yourself through the night is
sometimes an illusion. And despite the advertising, most people remain
skeptical. In a 2005 Gallup poll, more than 75% of people over 60 said
they were concerned about the long-term effects of sedative drugs, and
fewer than one in 10 regarded the drugs as “very safe.”
That doesn’t mean they are ineffective for everyone and in all
circumstances. For some chronically sleepless older people, a sedative
antidepressant or BRA may be almost a necessity. But most people should
use sleeping pills chiefly when they need uninterrupted sleep on a
special occasion or, at most, for a few weeks to prevent acute insomnia
from turning into chronic primary insomnia. Until we know more about the
long-term effects of eszopiclone and ramelteon, it cannot be said that
their advent changes the situation.
The most rigorously confirmed and widely recommended treatment for long-
term insomnia, especially in older people and those with psychiatric
disorders, includes several common features:
Stimulus control (reconditioning). A person who spends too much time
lying awake in bed (poor sleep efficiency) comes to associate the
bedroom with wakefulness. The problem is especially common in people
with a chronic physical or mental illness, and the solution is to break
the chain of association by reforming sleep habits. Patients are asked
to keep a diary recording those habits for a month or so to see which
thoughts and actions promote and interfere with a restful night. They
are encouraged to go to bed only when sleepy, avoid reading and watching
television in bed, and get up at the same time no matter how little they
have slept. They are also instructed to leave the bedroom after lying
awake for 20 minutes and return only when they feel sleepy again.
Most patients receiving stimulus- control therapy should avoid naps, but
that is not always advisable for older people because of their
relatively shallow and interrupted nighttime sleep. A regular nap of a
half-hour to two hours in the early afternoon may be helpful for them,
as long as the purpose is understood — not sleeping longer at night, but
being more active and clear-minded during the day.
Sleep restriction. This is an extension of stimulus control. The patient
goes to bed later than usual (so that she is more likely to fall asleep
immediately) but gets up at the usual time. Then bedtime is gradually
Sleep hygiene education. What you do in the daytime affects how you
sleep at night. The standard advice is to exercise regularly, but not in
the evening; get out into the sun if possible (most older people spend
more time indoors); don’t smoke; don’t eat a heavy meal shortly before
bedtime; avoid alcohol and caffeine; keep the bedroom dark and quiet,
using heavy curtains and earplugs or a white noise machine if necessary;
try a warm bath before going to bed.
Relaxation training. The body and mind can be prepared for sleep with
meditation, self-hypnosis, slow rhythmic breathing, or the repetition of
neutral words or visualization of soothing scenes. Another technique is
progressive muscle relaxation, which involves alternately tensing and
fully relaxing the muscles, starting with the feet. Older people can
skip the tensing if it is too difficult or uncomfortable.
Cognitive therapy. The aim is to correct unrealistic expectations of
perfect (or youthful) sleep and the catastrophic thinking about the
consequences of imperfect sleep that sometimes prevents relaxation and
keeps people awake. Cognitive therapy is usually part of a package that
includes behavioral treatments as well.
A behavioral sleep therapy has been developed by psychologists at
Flinders University in Adelaide, Australia. They have branded it
Flinders Accelerated Sleep Therapy (FAST), but it might also be called
Extreme Sleep Restriction. At 8 on a Saturday night, patients come to a
sleep laboratory, where they go to bed wired to an EEG machine that
tracks their brain waves. For 27 hours, until 11 p.m. Sunday, they are
awakened every time the EEG reading indicates that they’re falling
asleep — sometimes 50 times or more. In a preliminary study of patients
with severe chronic insomnia, just one round of FAST reduced sleep
latency (the time it took to fall asleep) by nearly 50%, from an hour
and 10 minutes to 40 minutes, and increased total sleep time from five
to six hours when measured six weeks later. The therapy is still
experimental, and there are no controlled studies.
Treating insomnia in psychiatric disorders
Insomnia in a person with depression or anxiety disorder is usually
called secondary insomnia, with the implication that it results from or
is a symptom of the anxiety or depression. Many used to think that
especially in older people, there was little point in treating insomnia
separately. Either it was considered a normal or inevitable effect of
age, or experts thought only treatment for the underlying psychiatric
disorder would help.
Opinion has changed. Many studies now show that behavioral treatment
specifically for sleep disturbances can be effective even in people
suffering from depression, anxiety, and posttraumatic stress disorder
(PTSD). For purposes of treatment, it may not be necessary to
distinguish between primary and secondary insomnia. A review found that
behavioral treatments for insomnia could be effectively combined with
psychotherapy for depression, anxiety, PTSD, obsessive-compulsive
disorder, and other psychiatric conditions. Good controlled studies are
in short supply, but in general, the results seem similar for primary
and secondary insomnia.
There is one caution: Sleep restriction can be risky for some people
with psychiatric disorders. It may heighten anxiety, trigger manic
episodes in patients with bipolar disorder, or raise the risk of daytime
panic attacks in people with panic disorder and agoraphobia.
Many experts are convinced that behavioral sleep therapy, although it
seems slower and more difficult, works better than drugs in the long run
for most patients. A controlled trial published in 2006, confirming
earlier research, found that a package of cognitive and behavioral
treatments was more effective than eszopiclone, the latest popular
sedative-hypnotic drug, in both the short term and the long term (six
months) for the treatment of primary insomnia in people with an average
age of 61. The drug had barely more effect than a placebo (sugar pill).
Patients given behavioral treatment spent less time lying awake in bed;
they had more deep sleep and their daytime functioning improved a little
more. Unfortunately, improvements are still modest compared to the
relief patients get from similar treatment for anxiety and depressions.
It’s not clear whether all parts of the cognitive behavioral therapy
package are actually needed. In 2006 a meta-analysis of 23 controlled
studies found that both stimulus control alone and relaxation training
alone were sufficient.
National Center on Sleep Disorders Research
National Sleep Foundation
National Institute on Aging Information Center
American Academy of Sleep Medicine
The future of sleep medicine
Sleep medicine researchers still have much to learn. They’re looking for
more selective drugs with fewer side effects, and ways to target
specific sleep problems — drugs that affect the sleep-wake cycle, brain
arousal circuits, the stress hormone system, and various
neurotransmitters and hormones involved in sleep regulation. It turns
out that one of these hormones may be orexin, now known chiefly as a
regulator of appetite.
Just as important, clinicians are increasingly treating insomnia as a
disorder as well as a symptom — a disorder that is often chronic,
especially in older people. The American Academy of Sleep Medicine has
started training programs in behavioral sleep medicine and is trying to
provide more access to behavioral treatments by encouraging group
therapy and the use of nurses as therapists. The hope is that as we gain
a better understanding and control of sleep, we will not have to resign
ourselves to accepting insomnia and its consequences as normal or
inevitable at any age.
Buysse DJ, ed. Sleep Disorders and Psychiatry. Review of Psychiatry,
Vol. 24. American Psychiatric Publications, 2005.
Glass J, et al. “Sedative-Hypnotics in Older People with Insomnia: Meta-
Analysis of Risks and Benefits,” BMJ (November 19, 2005): Vol. 331, pp.
Institute of Medicine, Board on Health Sciences Policy. Sleep Disorders
and Sleep Deprivation: An Unmet Public Health Problem. National
Academies Press, 2006.
Nau SD, et al. “Treatment of Insomnia in Older Adults,” Clinical
Psychology Review (July 2005): Vol. 25, No. 5, pp. 645-72.
Silber MH. “Chronic Insomnia,” New England Journal of Medicine (August
25, 2005): Vol. 353, No. 8, pp. 803-10.