Why Can’t Americans Sleep?
Insomnia has become a public-health
emergency.
By Jennifer Senior
June 30, 2025, 6 AM ET
Ilike to tell people that the night before I
stopped sleeping, I slept. Not only that: I slept well. Years ago, a boyfriend
of mine, even-keeled during the day but restless at night, told me how hard it
was to toss and turn while I instantly sank into the crude, Neanderthal slumber
of the dead. When I found a magazine job that allowed me to keep night-owl
hours, my rhythms had the precision of an atomic clock. I fell asleep at 1 a.m.
I woke up at 9 a.m. One to nine, one to nine, one to nine, night after night,
day after day. As most researchers can tell you, this click track is essential
to health outcomes: One needs consistent bedtimes and wake-up times. And I had
them, naturally; when I lost my alarm clock, I didn’t bother getting another
until I had an early-morning flight to catch.
Then, one night maybe two months
before I turned 29, that vaguening sense that normal sleepers have when they’re
lying in bed—their thoughts pixelating into surreal images, their mind listing
toward unconsciousness—completely deserted me. How bizarre, I
thought. I fell asleep at 5 a.m.
This started to happen pretty
frequently. I had no clue why. The circumstances of my life, both personally
and professionally, were no different from the week, month, or two months
before—and my life was good. Yet I’d somehow transformed into an appliance
without an off switch.
I saw an acupuncturist. I took Tylenol
PM. I sampled a variety of supplements, including melatonin (not really
appropriate, I’d later learn, especially in the megawatt doses Americans
take—its real value is in resetting your circadian clock, not as a sedative). I
ran four miles every day, did breathing exercises, listened to a meditation
tape a friend gave me. Useless.
I finally caved and saw my general
practitioner, who prescribed Ambien, telling me to feel no shame if I needed it
every now and then. But I did feel shame, lots of shame, and I’d always been
phobic about drugs, including recreational ones. And now … a sedative? (Two
words for you: Judy Garland.) It was only when I started enduring semiregular
involuntary all-nighters—which I knew were all-nighters, because I got out of
bed and sat upright through them, trying to read or watch TV—that I
capitulated. I couldn’t continue to stumble brokenly through the world after
nights of virtually no sleep.
I hated Ambien. One of the dangers
with this strange drug is that you may do freaky things at 4 a.m. without
remembering, like making a stack of peanut-butter sandwiches and eating them.
That didn’t happen to me (I don’t think?), but the drug made me squirrelly and
tearful. I stopped taking it. My sleep went back to its usual syncopated
disaster.
In Sleepless:
A Memoir of Insomnia, Marie Darrieussecq lists the thinkers and
artists who have pondered the brutality of sleeplessness, and they’re
distinguished company: Duras, Gide, Pavese, Sontag, Plath, Dostoyevsky,
Murakami, Borges, Kafka. (Especially Kafka, whom she calls literature’s “patron
saint” of insomniacs. “Dread of night,” he wrote. “Dread of not-night.”) Not to
mention F. Scott Fitzgerald, whose sleeplessness was triggered by a single
night of warfare with a mosquito.
But there was sadly no way to
interpret my sleeplessness as a nocturnal manifestation of tortured genius or
artistic brilliance. It felt as though I’d been poisoned. It was that
arbitrary, that abrupt. When my insomnia started, the experience wasn’t just
context-free; it was content-free. People would ask what I was thinking while
lying wide awake at 4 a.m., and my answer was: nothing. My mind whistled like a
conch shell.
But over time I did start thinking—or
worrying, I should say, and then perseverating, and then outright panicking. At
first, songs would whip through my head, and I couldn’t get the orchestra to
pack up and go home. Then I started to fear the evening, going to bed too early
in order to give myself extra runway to zonk out. (This, I now know, is a
typical amateur’s move and a horrible idea, because the bed transforms from a
zone of security into a zone of torment, and anyway, that’s not how the
circadian clock works.) Now I would have conscious thoughts
when I couldn’t fall asleep, which can basically be summarized as insomnia
math: Why am I not falling asleep Dear God let me fall asleep Oh my God
I only have four hours left to fall asleep oh my God now I only have three oh
my God now two oh my God now just one.
“The insomniac is not so much in
dialogue with sleep,” Darrieussecq writes, “as with the apocalypse.”
I would shortly discover that this
cycle was textbook insomnia perdition: a fear of sleep loss that itself causes sleep
loss that in turn generates an even greater fear of sleep loss
that in turn generates even more sleep loss … until the next
thing you know, you’re in an insomnia galaxy spiral, with a dark behavioral and
psychological (and sometimes neurobiological) life of its own.
I couldn’t recapture my nights.
Something that once came so naturally now seemed as impossible as flying. How
on earth could this have happened? To this day, whenever I think about it, I
still can’t believe it did.
In light of my tortured history with the
subject, you can perhaps see why I generally loathe stories about sleep. What
they’re usually about is the dangers of sleep loss, not sleep
itself, and as a now-inveterate insomniac, I’ve already got a multivolume
fright compendium in my head of all the terrible things that can happen when
sleep eludes you or you elude it. You will die of a heart attack or a stroke.
You will become cognitively compromised and possibly dement. Your weight will
climb, your mood will collapse, the ramparts of your immune system will
crumble. If you rely on medication for relief, you’re doing your disorder all
wrong—you’re getting the wrong kind of sleep, an unnatural sleep, and addiction
surely awaits; heaven help you and that horse of Xanax you rode in on.
It should go without saying that for
some of us, knowledge is not power. It’s just more kindling.
The cultural discussions around sleep
would be a lot easier if the tone weren’t quite so hectoring—or so smug. A case
in point: In 2019, the neuroscientist Matthew Walker, the author of Why
We Sleep, gave a TED Talk that began with a cheerful
disquisition about testicles. They are, apparently, “significantly smaller” in
men who sleep five hours a night rather than seven or more, and that two-hour
difference means lower testosterone levels too, equivalent to those of someone
10 years their senior. The consequences of short sleep for women’s reproductive
systems are similarly dire.
“This,” Walker says just 54 seconds
in, “is the best news that I have for you today.”
He makes good on his promise. What
follows is the old medley of familiars, with added verses about inflammation,
suicide, cancer. Walker’s sole recommendation at the end of his sermon is the
catechism that so many insomniacs—or casual media consumers, for that
matter—can recite: Sleep in a cool room, keep your bedtimes and wake-up times
regular, avoid alcohol and caffeine. Also, don’t nap.
I will now say about Walker:
1. His book is in many ways quite
wonderful—erudite and wide-ranging and written with a flaring energy when it
isn’t excessively pleased with itself.
2. Both Why We Sleep and
Walker’s TED Talk focus on sleep deprivation, not insomnia, with the implicit
and sometimes explicit assumption that too many people choose to blow off sleep
in favor of work or life’s various seductions.
If public awareness is Walker’s goal
(certainly a virtuous one), he and his fellow researchers have done a very good
job in recent years, with the enthusiastic assistance of my media colleagues,
who clearly find stories about the hazards of sleep deprivation irresistible.
(In the wine-dark sea of internet content, they’re click sirens.) Walker’s TED
Talk has been viewed nearly 24 million times. “For years, we were fighting
against ‘I’ll sleep when I’m dead,’ ” Aric Prather, the
director of the behavioral-sleep-medicine research program at UC San Francisco,
told me. “Now the messaging that sleep is a fundamental pillar of human health
has really sunk in.”
Yet greater awareness of sleep
deprivation’s consequences hasn’t translated into a better-rested populace.
Data from the CDC show that the proportion of Americans reporting insufficient
sleep held constant from 2013 through 2022, at roughly
35 percent. (From 2020 to 2022, as anxiety about the pandemic eased, the
percentage actually climbed.)
So here’s the first question I have:
In 2025, exactly how much of our “sleep opportunity,” as the experts call it,
is under our control?
According to the most recent
government data, 16.4 percent of American employees work
nonstandard hours. (Their health suffers in every category—the World
Health Organization now describes night-shift work as “probably carcinogenic.”) Adolescents live in a
perpetual smog of sleep deprivation because they’re forced to rise far too
early for school (researchers call their plight “social jet lag”); young
mothers and fathers live in a smog of sleep deprivation because they’re forced
to rise far too early (or erratically) for their kids; adults caring for aging
parents lose sleep too. The chronically ill frequently can’t sleep. Same with
some who suffer from mental illness, and many veterans, and many active-duty
military members, and menopausal women, and perimenopausal women, and the
elderly, the precariat, the poor.
“Sleep opportunity is not evenly
distributed across the population,” Prather noted, and he suspects that this
contributes to health disparities by class. In 2020, the National Center for
Health Statistics found that the poorer Americans were, the greater their
likelihood of reporting difficulty falling asleep. If you look at the CDC map
of the United States’ most sleep-deprived communities, you’ll see that
they loop straight through the Southeast and Appalachia. Black and Hispanic
Americans also consistently report sleeping less, especially Black women.
Even for people who aren’t contending
with certain immutables, the cadences of modern life have proved inimical to
sleep. Widespread electrification laid waste to our circadian rhythms 100 years
ago, when they lost any basic correspondence with the sun; now, compounding
matters, we’re contending with the currents of a wired world. For white-collar
professionals, it’s hard to imagine a job without the woodpecker incursions of
email or weekend and late-night work. It’s hard to imagine news consumption, or
even ordinary communication, without the overstimulating use of phones and
computers. It’s hard to imagine children eschewing social media when it’s how
so many of them socialize, often into the night, which means blue-light
exposure, which means the suppression of melatonin. (Melatonin suppression
obviously applies to adults too—it’s hardly like we’re avatars of discipline
when it comes to screen time in bed.)
Most of us can certainly do more to
improve or reclaim our sleep. But behavioral change is difficult, as anyone
who’s vowed to lose weight can attest. And when the conversation around sleep
shifts the onus to the individual—which, let’s face it, is the American way (we
shift the burden of child care to the individual, we shift the burden of health
care to the individual)—we sidestep the fact that the public and private
sectors alike are barely doing a thing to address what is essentially a
national health emergency.
Given that we’ve decided that an
adequate night’s rest is a matter of individual will, I now have a second
question: How are we to discuss those who are suffering not just from
inadequate sleep, but from something far more severe? Are we to lecture them in
the same menacing, moralizing way? If the burden of getting enough sleep is on
us, should we consider chronic insomniacs—for whom sleep is a nightly
gladiatorial struggle—the biggest failures in the armies of the underslept?
Those who can’t sleep suffer a great
deal more than those gifted with sleep will ever know. Yet insomniacs
frequently feel shame about the solutions they’ve sought for relief—namely,
medication—likely because they can detect a subtle, judgmental undertone about
this decision, even from their loved ones. Resorting to drugs means they are
lazy, refusing to do simple things that might ease their passage into
unconsciousness. It means they are neurotic, requiring pills to transport them
into a natural state that every other animal on Earth finds without aid.
Might I suggest that these views are
unenlightened? “In some respects, chronic insomnia is similar to where
depression was in the past. We’d say, ‘Major depression’ and people would say,
‘Everybody gets down now and then,’ ” John Winkelman, a
psychiatrist in the sleep-medicine division at Harvard Medical School, said at
a panel I attended last summer. Darrieussecq, the author of Sleepless, puts it more bluntly: “ ‘I didn’t sleep all night,’ sleepers say to
insomniacs, who feel like replying that they haven’t slept all their
life.”
The fact is, at least 12 percent of
the U.S. population suffers from insomnia as an obdurate condition.
Among Millennials, the number pops up to 15 percent. And 30 to 35 percent of
Americans suffer from some of insomnia’s various symptoms—trouble falling
asleep, trouble staying asleep, or waking too early—at least temporarily. In
2024, there were more than 2,500 sleep-disorder centers in the U.S. accredited
by the American Academy of Sleep Medicine. Prather told me the wait time to get
into his sleep clinic at UCSF is currently a year. “That’s better than it used
to be,” he added. “Until a few months ago, our waitlist was closed. We couldn’t
fathom giving someone a date.”
So what I’m hoping to do here is not
write yet another reproachful story about sleep, plump with misunderstandings
and myths. Fixing sleep—obtaining sleep—is a tricky business. The
work it involves and painful choices it entails deserve nuanced examination.
Contrary to what you might have read, our dreams are seldom in black and white.
Whenever I interviewed a clinician,
psychiatrist, neuroscientist, or any other kind of expert for this story, I
almost always opened with the same question: What dogma about sleep do you
think most deserves to be questioned?
The most frequent answer, by a long
chalk, is that we need eight hours of it. A fair number of studies, it turns
out, show that mortality rates are lowest if a person gets roughly seven hours.
Daniel F. Kripke, a psychiatrist at UC San Diego, published the most famous of
these analyses in 2002, parsing a sample of 1.1 million individuals and
concluding that those who reported more than eight hours of sleep a night
experienced significantly increased mortality rates. According to
Kripke’s work, the optimal sleep range was a mere 6.5 to 7.4 hours.
These numbers shouldn’t be taken as
gospel. The relationship between sleep duration and health outcomes is a
devil’s knot, though Kripke did his best to control for the usual
confounds—age, sex, body-mass index. But he could not control for the factors he
did not know. Perhaps many of the individuals who slept eight hours or more
were doing so because they had an undetected illness, or an illness of greater
severity than they’d realized, or other conditions Kripke hadn’t accounted for.
The study was also observational, not randomized.
But even if they don’t buy Kripke’s
data, sleep experts don’t necessarily believe that eight hours of sleep has
some kind of mystical significance. Methodologically speaking, it’s hard to
determine how much sleep, on average, best suits us, and let’s not forget the
obvious: Sleep needs—and abilities—vary over the course of a lifetime, and from
individual to individual. (There’s even an extremely rare species of people,
known as “natural short sleepers,” associated with a handful of genes, who
require only four to six hours a night. They tear through the world as if fired
from a cannon.) Yet eight hours of sleep or else remains one
of our culture’s most stubborn shibboleths, and an utter tyranny for many
adults, particularly older ones.
“We have people coming into our
insomnia clinic saying ‘I’m not sleeping eight hours’ when they’re 70 years of
age,” Michael R. Irwin, a psychoneurologist at UCLA, told me. “And the average
sleep in that population is less than seven hours. They attribute all kinds of
things to an absence of sleep—decrements in cognitive performance and vitality,
higher levels of fatigue—when often that’s not the case. I mean, people get
older, and the drive to sleep decreases as people age.”
Another declaration I was delighted to
hear: The tips one commonly reads to get better sleep are as insipid as they
sound. “Making sure that your bedroom is cool and comfortable, your bed is
soft, you have a new mattress and a nice pillow—it’s unusual that those things
are really the culprit,” Eric Nofzinger, the former director of the sleep
neuroimaging program at the University of Pittsburgh’s medical school, told me.
“Most people self-regulate anyway. If they’re cold, they put on an extra
blanket. If they’re too warm, they throw off the blanket.”
“Truthfully, there’s not a lot of data
supporting those tips,” Suzanne Bertisch, a behavioral-sleep-medicine expert at
Brigham and Women’s Hospital, in Boston, told me. That includes the
proscription on naps, she added, quite commonly issued in her world. (In
general, the research on naps suggests that short ones have beneficial outcomes
and long ones have negative outcomes, but as always, cause and effect are
difficult to disentangle: An underlying health condition could be driving those
long naps.)
Even when they weren’t deliberately
debunking the conventional wisdom about sleep, many of the scholars I spoke
with mentioned—sometimes practically as an aside—facts that surprised or
calmed. For instance: Many of us night owls have heard that the weather
forecast for our old age is … well, cloudy, to be honest, with a late-afternoon
chance of keeling over. According to one large analysis, we have a 10 percent
increase in all-cause mortality over morning larks. But Jeanne Duffy, a
neuroscientist distinguished for her expertise in human circadian rhythms at
Brigham and Women’s, told me she suspected that this was mainly because most
night owls, like most people, are obliged to rise early for their job.
So wait, I said. Was she implying that
if night owls could contrive work-arounds to suit their biological inclination
to go to bed late, the news probably wouldn’t be as grim?
“Yes,” she replied.
A subsequent study showed that the
owl-lark mortality differential dwindled to nil when the authors controlled for lifestyle.
Apparently owls are more apt to smoke, and to drink more. So if you’re an owl
who’s repelled by Marlboros and Jameson, you’re fine.
Kelly Glazer Baron, the director of
the behavioral-sleep-medicine program at the University of Utah, told me that
she’d love it if patients stopped agonizing over the length of their individual
sleep phases. I didn’t get enough deep sleep, they fret, thrusting
their Apple Watch at her. I didn’t get enough REM. And yes,
she said, insufficiencies in REM or slow-wave sleep can be a problem,
especially if they reflect an underlying health issue. But clinics don’t look
solely at sleep architecture when evaluating their patients.
“I often will show them my own data,”
Baron said. “It always shows I don’t have that much deep
sleep, which I find so weird, because I’m a healthy middle-aged woman.” In
2017, after observing these anxieties for years, Baron coined a term for sleep
neuroticism brought about by wearables: orthosomnia.
But most surprising—to me, anyway—was
what I heard about insomnia and the black dog. “There are far more studies
indicating that insomnia causes depression than depression causes insomnia,”
said Wilfred Pigeon, the director of the Sleep & Neurophysiology Research
Laboratory at the University of Rochester. Which is not to say, he added, that
depression can’t or doesn’t cause insomnia. These forces, in the parlance of
health professionals, tend to be “bidirectional.”
But I can’t tell you how vindicating I
found the idea that perhaps my own insomnia came first. A couple of years into
my struggles with sleeplessness, a brilliant psychopharmacologist told me that
my new condition had to be an episode of depression in disguise. And part of me
thought, Sure, why not? A soundtrack of melancholy had been
playing at a low hum inside my head from the time I was 10.
The thing was: I became outrageously
depressed only after my insomnia began. That’s when that low hum
started to blare at a higher volume. Until I stopped sleeping, I never suffered
from any sadness so crippling that it prevented me from experiencing joy. It
never impeded my ability to socialize or travel. It never once made me
contemplate antidepressants. And it most certainly never got in the way of my
sleeping. The precipitating factor in my own brutal insomnia was, and remains,
an infuriating mystery.
Sleep professionals, I have learned, drink a lot of
coffee. That was the first thing I noticed when I attended SLEEP 2024, the
annual conference of the American Academy of Sleep Medicine, in Houston:
coffee, oceans of it, spilling from silver urns, especially at the industry
trade show. Wandering through it was a dizzying experience, a sprawling
testament to the scale and skyscraping profit margins of Big Sleep. More than
150 exhibitors showed up. Sheep swag abounded. Drug reps were everywhere, their
aggression tautly disguised behind android smiles, the meds they hawked called
the usual names that look like high-value Scrabble words.
I’ve never understood this branding
strategy, honestly. If you want your customers to believe they’re falling into
a gentle, natural sleep, you should probably think twice before calling your
drug Quviviq
I walked through the cavernous hall in
a daze. It was overwhelming, really—the spidery gizmos affixed to armies of
mannequins, the Times Square–style digital billboards screaming about the
latest in sleep technology.
At some point it occurred to me that
the noisy, overbusy, fluorescent quality of this product spectacular reminded
me of the last place on Earth a person with a sleep disorder should be: a
casino. The room was practically sunless. I saw very few clocks. After I spent
an afternoon there, my circadian rhythms were shot to hell.
But the conference itself …!
Extraordinary, covering miles of ground. I went to one symposium about “sleep
deserts,” another about the genetics of sleep disturbance, and yet another
about sleep and menopause. I walked into a colloquy about sleep and screens and
had to take a seat on the floor because the room was bursting like a suitcase.
Of most interest to me, though, were two panels, which I’ll shortly discuss:
one about how to treat patients with anxiety from new-onset insomnia, and one
on whether hypnotics are addictive.
My final stop at the trade fair was
the alley of beauty products—relevant, I presume, because they address the
aesthetic toll of sleep deprivation. Within five minutes, an energetic young
salesman made a beeline for me, clearly having noticed that I was a woman of a
certain age. He gushed about a $2,500 infrared laser to goose collagen
production and a $199 medical-grade peptide serum that ordinarily retails for
$1,100. I told him I’d try the serum. “Cheaper than Botox, and it does the same
thing,” he said approvingly, applying it to the crow’s-feet around my eyes.
I stared in the mirror. Holy shit. The
stuff was amazing.
“I’ll take it,” I told him.
He was delighted. He handed me a box.
The serum came in a gold syringe.
“You’re a doctor, right?”
A beat.
“No,” I finally said. “A journalist.
Can only a dermatologist—”
He told me it was fine; it’s just that
doctors were his main customers. This was the sort of product women like me
usually had to get from them. I walked away elated but queasy, feeling like a
creep who’d evaded a background check by purchasing a Glock at a gun show.
The first line of treatment for chronic,
intractable sleeplessness, per the American Academy of Sleep Medicine, is
cognitive behavioral therapy for insomnia, or CBT-I. I’ve tried it, in earnest,
at two different points in my life. It generally involves six to eight sessions
and includes, at minimum: identifying the patient’s sleep-wake patterns
(through charts, diaries, wearables); “stimulus control” (setting consistent
bedtimes and wake-up times, resisting the urge to stare at the clock, delinking
the bed from anything other than sleep and sex); establishing good sleep habits
(the stuff of every listicle); “sleep restriction” (compressing your sleep
schedule, then slowly expanding it over time); and “cognitive restructuring,”
or changing unhealthy thoughts about sleep.
The cognitive-restructuring component
is the most psychologically paradoxical. It means taking every terrifying thing
you’ve ever learned about the consequences of sleeplessness and pretending
you’ve never heard them.
I pointed this out to Wilfred Pigeon.
“For the medically anxious, it’s tough,” he agreed. “We’re trying to tell
patients two things at the same time: ‘You really need to get your sleep on
track, or you will have a heart attack five years earlier than you otherwise
would.’ But also: ‘Stop worrying about your sleep so much, because it’s
contributing to your not being able to sleep.’ And they’re both true!”
Okay, I said. But if an insomniac
crawls into your clinic after many years of not sleeping (he says people tend
to wait about a decade), wouldn’t they immediately see that these two messages
live in tension with each other? And dwell only on the heart attack?
“I tell the patient their past
insomnia is water under the bridge,” Pigeon said. “We’re trying to erase
the added risks that ongoing chronic insomnia will have. Just
because a person has smoked for 20 years doesn’t mean they should keep
smoking.”
He’s absolutely right. But I’m not
entirely convinced that these incentives make the cognitive dissonance of CBT-I
go away. When Sara Nowakowski, a CBT-I specialist at Baylor College of
Medicine, gave her presentation at SLEEP 2024’s panel on anxiety and new-onset
insomnia, she said that many of her patients start reciting the grim data from
their Fitbits and talking about dementia.
That’s likely because they’ve read the
studies. Rapid-eye-movement (REM) sleep, that vivid-dream stage when our eyes
race beneath our eyelids like mice under a blanket, is essential to emotional
regulation and problem-solving. Slow-wave sleep, our deepest sleep, is
essential for repairing our cells, shoring up our immune systems, and rinsing
toxins from our brains, thanks to a watery complex of micro-canals called the
glymphatic system. We repair our muscles when we sleep. We restore our hearts.
We consolidate memories and process knowledge, embedding important facts and
disposing of trivial ones. We actually learn when we’re
asleep.
Many insomniacs know all too well how
nonnegotiably vital sleep is, and what the disastrous consequences are if you
don’t get it. I think of the daredevil experiment that Nathaniel Kleitman, the
father of sleep research, informally conducted as a graduate student in 1922,
enlisting five classmates to join him in seeing how long they could stay awake.
He lasted the longest—a staggering 115 hours—but at a terrible price,
temporarily going mad with exhaustion, arguing on the fifth day with an
imaginary foe about the need for organized labor. And I think of Allan
Rechtschaffen, another pioneer in the field, who in 1989 had the fiendish idea
to place rats on a spinning mechanism that forced them to stay
awake if they didn’t want to drown. They eventually dropped dead.
So these are the kinds of facts a
person doing CBT-I has to ignore.
Still. Whether a patient’s terrors concern the
present or the future, it is the job of any good CBT-I practitioner to help
fact-check or right-size them through Socratic questioning. During her panel at
SLEEP 2024, Nowakowski gave very relatable examples:
When you’re struggling to fall asleep,
what are you most worried will happen?
I’ll lose my job/scream at my
kids/detonate my relationship/never be able to sleep again.
And what’s the probability of your not
falling asleep?
I don’t sleep most nights.
And the probability of not functioning
at work or yelling at the kids if you don’t?
Ninety percent.
She then tells her patients to go read
their own sleep diary, which she’s instructed them to keep from the start. The
numbers seldom confirm they’re right, because humans are monsters of
misprediction. Her job is to get her patients to start decatastrophizing, which
includes what she calls the “So what?” method: So what if you have a
bad day at work or at home? You’ve had others. Will it be the end of the world? (When
my second CBT-I therapist asked me this, I silently thought, Yes,
because when I’m dangling at the end of my rope, I just spin more.) CBT-I
addresses anxiety about not sleeping, which tends to be the real
force that keeps insomnia airborne, regardless of what lofted it. The pre-sleep
freaking out, the compulsive clock-watching, the bargaining, the
middle-of-the-night doom-prophesizing, the despairing—CBT-I attempts to snip
that loop. The patient actively learns new behaviors and attitudes to put an
end to their misery.
But the main anchor of CBT-I is
sleep-restriction therapy. I tried it back when I was 29, when I dragged my
wasted self into a sleep clinic in New York; I’ve tried it once since. I
couldn’t stick with it either time.
The concept is simple: You severely
limit your time in bed, paring away every fretful, superfluous minute you’d
otherwise be awake. If you discover from a week’s worth of sleep-diary entries
(or your wearable) that you spend eight hours buried in your duvet but sleep
for only five of them, you consolidate those splintered hours into one bloc of
five, setting the same wake-up time every day and going to bed a mere five
hours before. Once you’ve averaged sleeping those five hours for a few days
straight, you reward your body by going to bed 15 minutes earlier. If you
achieve success for a few days more, you add another 15 minutes. And then
another … until you’re up to whatever the magic number is for you.
No napping. The idea is to build up
enough “sleep pressure” to force your body to collapse in surrender.
Sleep restriction can be a wonderful
method. But if you have severe insomnia, the idea of reducing your sleep time
is petrifying. Technically, I suppose, you’re not really reducing
your sleep time; you’re just consolidating it. But practically speaking, you
are reducing your sleep, at least in the beginning, because dysregulated sleep
isn’t an accordion, obligingly contracting itself into a case. Contracting it
takes time, or at least it did for me. The process was murder.
“If you get people to really work
their way through it—and sometimes that takes holding people’s hands—it ends up
being more effective than a pill,” Ronald Kessler, a renowned psychiatric
epidemiologist at Harvard, told me when I asked him about CBT-I. The problem is
the formidable size of that if. “CBT-I takes a lot more work than
taking a pill. So a lot of people drop out.”
They do. One study I perused had an
attrition rate of 40 percent.
Twenty-six years ago, I, too, joined
the legions of the quitters. In hindsight, my error was my insistence on trying
this grueling regimen without a benzodiazepine (Valium, Ativan, Xanax), though
my doctor had recommended that I start one. But I was still afraid of drugs in
those days, and I was still in denial that I’d become hostage to my own brain’s
terrorism. I was sure that I still had the power to negotiate. Competence had
until that moment defined my whole life. I persuaded the doctor to let me try without
drugs.
As she’d predicted, I failed. The
graphs in my sleep diary looked like volatile weeks on the stock exchange.
For the first time ever, I did need an
antidepressant. The doctor wrote me a prescription for Paxil and a bottle of
Xanax to use until I got up to cruising altitude—all SSRIs take a while to kick
in.
I didn’t try sleep restriction again
until many years later. Paxil sufficed during that time; it made me almost
stupid with drowsiness. I was sleepy at night and vague during the day. I
needed Xanax for only a couple of weeks, which was just as well, because I
didn’t much care for it. The doctor had prescribed too powerful a dose, though
it was the smallest one. I was such a rookie with drugs in those days that it
never occurred to me I could just snap the pill in half.
Have I oversimplified the story of my
insomnia? Probably. At the top of the SLEEP 2024 panel about anxiety and
new-onset insomnia, Leisha Cuddihy, a director at large for the Society of
Behavioral Sleep Medicine, said something that made me wince—namely, that her
patients “have a very vivid perception of pre-insomnia sleep being literally
perfect: ‘I’ve never had a bad night of sleep before now.’ ”
Okay, guilty as charged. While it’s
true that I’d slept brilliantly (and I stand by this, brilliantly)
in the 16 years before I first sought help, I was the last kid to fall asleep
at slumber parties when I was little. Cuddihy also said that many of her
patients declare they’re certain, implacably certain, that they are unfixable.
“They feel like something broke,” she said.
Which is what I wrote just a few pages
back. Poisoned, broke, same thing.
By the time Cuddihy finished speaking,
I had to face an uncomfortable truth: I was a standard-issue sleep-clinic
zombie.
But when patients say they feel like
something broke inside their head, they aren’t necessarily wrong. An
insomniac’s brain does change in neurobiological ways.
“There is something in the neurons
that’s changing during sleep in patients with significant sleep disruptions,”
said Eric Nofzinger, who, while at the University of Pittsburgh, had one of the
world’s largest databases of brain-imaging studies of sleeping human beings.
“If you’re laying down a memory, then that circuitry is hardwired for that
memory. So one can imagine that if your brain is doing this night after night
…”
We know that the
hypothalamic-pituitary-adrenal axis, our body’s first responder to stress, is
overactivated in the chronically underslept. If the insomniac suffers from
depression, their REM phase tends to be longer and more “dense,” with the
limbic system (the amygdala, the hippocampus—where our primal drives are
housed) going wild, roaring its terrible roars and gnashing its terrible teeth.
(You can imagine how this would also make depressives subconsciously less
motivated to sleep—who wants to face their Gorgon dreams?) Insomniacs suffering
from anxiety experience this problem too, though to a lesser degree; it’s their
deep sleep that’s mainly affected, slimming down and shallowing out.
And in all insomniacs, throughout the
night, the arousal centers of the brain keep clattering away, as does the
prefrontal cortex (in charge of planning, decision making), whereas in regular
sleepers, these buzzing regions go offline. “So when someone with insomnia
wakes up the next morning and says, ‘I don’t think I slept at all last night,’
in some respects, that’s true,” Nofzinger told me. “Because the parts of the
brain that should have been resting did not.”
And why didn’t they rest? The
insomniac can’t say. The insomniac feels at once responsible and helpless when
it comes to their misery: I must be to blame. But I can’t be to blame. The
feeling that sleeplessness is happening to you, not something you’re doing to
yourself, sends you on a quest for nonpsychological explanations: Lots of
physiological conditions can cause sleep disturbances, can’t they? Obstructive
sleep apnea, for instance, which afflicts nearly 30 million Americans. Many
autoimmune diseases, too. At one point, I’ll confess that I started asking the
researchers I spoke with whether insomnia itself could be an autoimmune
disorder, because that’s what it feels like to me—as if my brain is going after
itself with brickbats.
“Narcolepsy appears to be an example
of a sleep disorder involving the immune system,” Andrew Krystal, a
psychiatrist specializing in sleep disorders at UCSF, told me.
What? I said. Really?
Really, he replied. “There are few
things I know of,” he said, “that are as complicated as the mammalian immune
system.”
But insomnia-as-autoimmune-disorder is
only a wisp of a theory, a wish of a theory, nothing more. In her memoir, The
Shapeless Unease: A Year of Not Sleeping, the novelist Samantha
Harvey casts around for a physiological explanation, too. But after she
completes a battery of tests, the results come back normal, pointing to “what I
already know,” she writes, “which is that my sleeplessness is psychological. I
must carry on being the archaeologist of myself, digging around, seeing if I
can excavate the problem and with it the solution—when in truth I am afraid of
myself, not of what I might uncover, but of managing to uncover nothing.”
I didn’t tolerate my Paxil brain for long. I
weaned myself off, returned to normal for a few months, and assumed that my
sleeplessness had been a freak event, like one of those earthquakes in a city
that never has them. But then my sleep started to slip away again, and by age
31, I couldn’t recapture it without chemical assistance. Prozac worked for
years on its own, but it blew out whatever circuit in my brain generates
metaphors. When I turned to the antidepressants that kept the electricity
flowing, I needed sleep medication too—proving, to my mind, that melancholy
couldn’t have been the mother of my sleep troubles, but the lasting result of
them. I’ve used the lowest dose of Klonopin to complement my SSRIs for years.
In times of acute stress, I need a gabapentin or a Unisom too.
Unisom is fine. Gabapentin also turns
my mind into an empty prairie.
Edibles, which I’ve also tried, turn
my brain to porridge the next day. Some evidence suggests that cannabis works
as a sleep aid, but more research, evidently, is required. (Sorry.)
Which brings me to the subject of
drugs. I come neither to praise nor to bury them. But I do come to reframe the
discussion around them, inspired by what a number of researcher-clinicians said
about hypnotics and addiction during the SLEEP 2024 panel on the subject. They
started with a simple question: How do you define addiction?
It’s true that many of the people who
have taken sleep medications for months or years rely on them. Without them,
the majority wouldn’t sleep, at least in the beginning, and a good many would
experience rebound insomnia if they didn’t wean properly, which can be even
worse. One could argue that this dependence is tantamount to addiction.
But: We don’t say people are addicted
to their hypertension medication or statins, though we know that in certain
instances lifestyle changes could obviate the need for either one. We don’t say
people are addicted to their miracle GLP-1 agonists just because they could
theoretically diet and exercise to lose weight. We agree that they need them.
They’re on Lasix. On Lipitor. On Ozempic.
Not addicted to.
Yet we still think of sleep
medications as “drugs,” a word that in this case carries a whiff of
stigma—partly because mental illness still carries a stigma, but also because
sleep medications legitimately do have the potential for recreational use and
abuse.
But is that what most people who
suffer from sleep troubles are doing? Using their Sonata or Ativan for fun?
“If you see a patient who’s been
taking medication for a long time,” Tom Roth, the founder of the Sleep
Disorders and Research Center at Henry Ford Hospital, said during the panel,
“you have to think, ‘Are they drug-seeking or therapy-seeking ?’ ” The overwhelming
majority, he and other panelists noted, are taking their prescription drugs for
relief, not kicks. They may depend on them, but they’re not abusing them—by taking them during the
day, say, or for purposes other than sleep.
Still, let’s posit that many long-term
users of sleep medication do become dependent. Now let’s consider another
phenomenon commonly associated with reliance on sleep meds: You enter Garland
and Hendrix territory in a hurry. First you need one pill, then you need two;
eventually you need a fistful with a fifth of gin.
Yet a 2024 cohort study, which
involved nearly 1 million Danes who used benzodiazepines long-term, found that
of those who used them for three years or more—67,398 people, to be exact—only
7 percent exceeded their recommended dose.
Not a trivial number, certainly, if
you’re staring across an entire population. But if you’re evaluating the risk
of taking a hypnotic as an individual, you’d be correct to assume that your
odds of dose escalation are pretty low.
That there’s a difference between
abuse and dependence, that dependence doesn’t mean a mad chase for more
milligrams, that people depend on drugs for a variety of other naturally
reversible conditions and don’t suffer any stigma—these nuances matter.
“Using something where the benefits
outweigh the side effects certainly is not addiction,” Winkelman, the Harvard
psychiatrist and chair of the panel, told me when we spoke a few months later.
“I call that treatment.”
The problem, he told me, is when the
benefits stop outweighing the downsides. “Let’s say the medication loses
efficacy over time.” Right. That 7 percent. And over-the-counter sleep meds,
whose active component is usually diphenhydramine (more commonly known as
Benadryl), are potentially even more likely to lose their efficacy—the American
Academy of Sleep Medicine advises against them. “And let’s say you did stop
your medication,” Winkelman continued. “Your sleep could be worse than it was
before you started it,” at least for a while. “People should know about that
risk.”
A small but even more hazardous risk:
a seizure, for those who abruptly stop taking high doses of benzodiazepines
after they’ve been on them for a long period of time. The likelihood is low—the
exact percentage is almost impossible to ascertain—but any risk of a seizure is
worth knowing about. “And are you comfortable with the idea that the drug could
irrevocably be changing your brain?” Winkelman asked. “The brain is a machine,
and you’re exposing it to the repetitive stimulus of the drug.” Then again, he pointed
out, you know what else is a repetitive stimulus? Insomnia.
“So should these things even be
considered a part of an addiction?” he asked. “At what point does a treatment
become an addiction? I don’t know.”
Calvinist about sleep meds, blasé about sleep
meds—whatever you are, the fact remains: We’re a nation that likes them.
According to a 2020 report from the National Center for Health Statistics, 8.4
percent of Americans take sleep medications most nights or every night, and an
additional 10 percent take them on some. Part of the reason medication remains
so popular is that it’s easy for doctors to prescribe a pill and give a patient
immediate relief, which is often what patients are looking for, especially if
they’re in extremis or need some assistance through a rough stretch. CBT‑I, as
Ronald Kessler noted, takes time to work. Pills don’t.
But another reason, as Suzanne
Bertisch pointed out during the addiction-and-insomnia-meds panel, is that
“primary-care physicians don’t even know what CBT-I is. This is a failure of
our field.”
Even if general practitioners did know
about CBT-I, too few therapists are trained in it, and those who are tend to
have fully saturated schedules. The military, unsurprisingly, has tried to work
around this problem (sleep being crucial to soldiers, sedatives being
contraindicated in warfare) with CBT-I via video as well as an online program,
both shown to be efficacious. But most of us are not in the Army. And while
some hospitals, private companies, and the military have developed apps for
CBT-I too, most people don’t know about them.
For years, medication has worked for
me. I’ve stopped beating myself up about it. If the only side effect I’m
experiencing from taking 0.5 milligrams of Klonopin is being dependent on 0.5
milligrams of Klonopin, is that really such a problem?
There’s been a lot of confusing noise
about sleep medication over the years. “Weak science, alarming FDA black-box
warnings, and media reporting have fueled an anti-benzodiazepine movement,”
says an editorial in the March 2024 issue of The American
Journal of Psychiatry. “This has created an atmosphere of fear
and stigma among patients, many of whom can benefit from such medications.”
A case in point: For a long time, the
public believed that benzodiazepines dramatically increased the risk of
Alzheimer’s disease, thanks to a 2014 study in the British Medical
Journal that got the usual five-alarm-fire treatment by the media.
Then, two years later, another study came along, also in the British
Medical Journal, saying, Never mind, nothing to see here, folks;
there appears to be no causal relationship we can discern.
That study may one day prove
problematic, too. But the point is: More work needs to be done.
A different paper, however—again by
Daniel Kripke, the fellow who argued that seven hours of sleep a night
predicted the best health outcomes—may provide more reason for concern. In a
study published in 2012, he looked at more than 10,000 people on a variety of
sleep medications and found that they were several times more likely to die
within 2.5 years than a matched cohort, even when controlling for a range of
culprits: age, sex, alcohol use, smoking status, body-mass index, prior cancer.
Those who took as few as 18 pills a year had a 3.6-fold increase. (Those who
took more than 132 had a 5.3-fold one.)
John Winkelman doesn’t buy it.
“Really,” he told me, “what makes a lot more sense is to ask, ‘Why did people
take these medications in the first place?’ ” And for what it’s worth, a 2023 study
funded by the National Institute on Drug Abuse and published in the Journal of the American Medical Association found that people on stable,
long-term doses of a benzodiazepine who go off their medication have worse
mortality rates in the following 12 months than those who stay on it. So maybe
you’re damned if you do, damned if you don’t.
Still, I take Kripke’s study
seriously. Because … well, Christ, I don’t know. Emotional reasons? Because
other esteemed thinkers still think there’s something to it?
In my own case, the most compelling
reasons to get off medication are the more mundane ones: the scratchy little
cognitive impairments it can cause during the day, the risk of falls as you get
older. (I should correct myself here: Falling when you’re older has the
potential to be not mundane, but very bad.) Medications can also cause problems
with memory as one ages, even if they don’t cause Alzheimer’s, and the
garden-variety brain termites of middle and old age are bummer enough.
And maybe most generally: Why have a
drug in your system if you can learn to live without it?
My suspicion is that most people who
rely on sleep drugs would prefer natural sleep.
So yes: I’d love to one day make a
third run at CBT-I, with the hope of weaning off my medication, even if it
means going through a hell spell of double exhaustion. CBT-I is a skill,
something I could hopefully deploy for the rest of my life. Something I can’t
accidentally leave on my bedside table.
Some part of me, the one that’s made
of pessimism, is convinced that it won’t work no matter how long I stick with
it. But Michael Irwin, at UCLA, told me something reassuring: His research
suggests that if you have trouble with insomnia or difficulty maintaining your
sleep, mindfulness meditation while lying in bed can be just as effective as
climbing out of bed, sitting in a chair, and waiting until you’re tired enough
to crawl back in—a pillar of CBT‑I, and one that I absolutely despise. I do it
sometimes, because I know I should, but it’s lonely and freezing, a form of
banishment.
And if CBT-I doesn’t work, Michael
Grandner, the director of the sleep-and-health-research program at the
University of Arizona, laid out an alternative at SLEEP 2024: acceptance and
commitment therapy, or ACT. The basic idea is exactly what the name suggests.
You accept your lot. You change exactly nothing. If you can’t sleep, or you
can’t sleep enough, or you can sleep only in a broken line, you say, This
is one of those things I can’t control. (One could see how such a mantra
might help a person sleep, paradoxically.) You then isolate what matters to
you. Being functional the next day? Being a good parent? A good friend? If
sleep is the metaphorical wall you keep ramming your head against, “is your
problem the wall?” Grandner asked. “Or is your problem that you can’t get
beyond the wall, and is there another way?”
Because there often is another way. To
be a good friend, to be a good parent, to be who and whatever it is you most
value—you can live out a lot of those values without adequate sleep. “When you
look at some of these things,” Grandner said, “what you find is that the
pain”—of not sleeping—“is actually only a small part of what is getting in the
way of your life. It’s really less about the pain itself and more about the
suffering around the pain, and that’s what we can fix.”
Even as I type, I’m skeptical of this
method too. My insomnia was so extreme at 29, and still can be to this day,
that I’m not sure I am tough enough—or can summon enough of my inner Buddha
(barely locatable on the best of days)—to transcend its pain, at once towering
and a bore. But if ACT doesn’t work, and if CBT-I doesn’t work, and if
mindfully meditating and acupuncture and neurofeedback and the zillions of
other things I’ve tried in the past don’t work on their own … well … I’ll go
back on medication.
Some people will judge me, I’m sure.
What can I say? It’s my life, not theirs.
I’ll wrap up by talking about an extraordinary
man named Thomas Wehr, once the chief of clinical psychobiology at the National
Institute of Mental Health, now 83, still doing research. He was by far the
most philosophical expert I spoke with, quick to find (and mull) the underlayer
of whatever he was exploring. I really liked what he had to say about sleep.
You’ve probably read the theory somewhere—it’s a media
chestnut—that human beings aren’t necessarily meant to sleep in one long
stretch but rather in two shorter ones, with a dreamy, middle-of-the-night
entr’acte. In a famous 2001 paper, the historian A. Roger Ekirch showed that
people in the pre-electrified British Isles used that interregnum to read,
chat, poke the fire, pray, have sex. But it was Wehr who, nearly 10 years
earlier, found a biological basis for these rhythms of social life, discovering
segmented sleep patterns in an experiment that exposed its participants to 14
hours of darkness each night. Their sleep split in two.
Wehr now knows firsthand what it is to
sleep a divided sleep. “I think what happens as you get older,” he told me last
summer, “is that this natural pattern of human sleep starts intruding back into
the world in which it’s not welcome—the world we’ve created with artificial
light.”
There’s a melancholy quality to this
observation, I know. But also a beauty: Consciously or not, Wehr is reframing
old age as a time of reintegration, not disintegration, a time when our natural
bias for segmented sleep reasserts itself as our lives are winding down.
His findings should actually be
reassuring to everyone. People of all ages pop awake in the middle of the night
and have trouble going back to sleep. One associates this phenomenon with
anxiety if it happens in younger people, and no doubt that’s frequently the
cause. But it also rhymes with what may be a natural pattern. Perhaps we’re
meant to wake up. Perhaps broken sleep doesn’t mean our sleep is broken,
because another sleep awaits.
And if we think of those
middle-of-the-night awakenings as meant to be, Wehr told me, perhaps we should
use them differently, as some of our forebears did when they’d wake up in the
night bathed in prolactin, a hormone that kept them relaxed and serene. “They
were kind of in an altered state, maybe a third state of consciousness you
usually don’t experience in modern life, unless you’re a meditator. And they
would contemplate their dreams.”
Night awakenings, he went on to
explain, tend to happen as we’re exiting a REM cycle, when our dreams are most
intense. “We’re not having an experience that a lot of our ancestors had of
waking up and maybe processing, or musing, or let’s even say ‘being informed’
by dreams.”
We should reclaim those moments at 3
or 4 a.m., was his view. Why not luxuriate in our dreams? “If you know you’re
going to fall back asleep,” he said, “and if you just relax and maybe think
about your dreams, that helps a lot.”
This assumes one has pleasant or
emotionally neutral dreams, of course. But I take his point. He was possibly
explaining, unwittingly, something about his own associative habits of
mind—that maybe his daytime thinking is informed by the meandering stories he
tells himself while he sleeps.
The problem, unfortunately, is that
the world isn’t structured to accommodate a second sleep or a day informed by
dreams. We live unnatural, anxious lives. Every morning, we turn on our lights,
switch on our computers, grab our phones; the whir begins. For now, this
strange way of being is exclusively on us to adapt to. Sleep doesn’t much curve
to it, nor it to sleep. For those who struggle each night (or day), praying for
what should be their biologically given reprieve from the chaos, the world has
proved an even harsher place.
But there are ways to improve it.
Through policy, by refraining from judgment—of others, but also of ourselves.
Meanwhile, I take comfort in the two hunter-gatherer tribes Wehr told me about,
ones he modestly noted did not confirm his hypothesis of
biphasic sleep. He couldn’t remember their names, but I later looked them up:
the San in Namibia and the Tsimané in Bolivia. They average less than 6.5 hours
of sleep a night. And neither has a word for insomnia.
This article appears in the August 2025 print edition with the headline
“American Insomnia.”
