This story is from The Pulse, a weekly health and science podcast.
Updated 2:45 p.m. July 29.
For Michael, getting enough sleep isn’t just an important part of staying healthy. It’s a matter of life or death.
“Really, the whole sleep-wake cycle and the amount of sleep and things is so pivotal,” he said. “I mean, it’s really a starting point for everything.”
Michael’s troubles started in his early 30s, when he hit a streak of bad luck. He got divorced, was laid off from his job, then got into a serious car accident that resulted in a severe concussion. (He spoke on the condition that only his first name be used.)
“It was a big head injury,” Michael said. “And after that, I started not being able to sleep.”
It was a gradual slide, at first. With no family or job to anchor his schedule, Michael began sleeping later and later. After the divorce, he was forced to move from his family’s waterfront home in the Seattle area into a cramped trailer that belonged to a friend.
“I had no money,” he said. “I was unemployed … It was just squalid conditions, just absolutely nothing.”
The trailer got almost zero sunlight, and soon Michael’s sleep schedule had gone completely haywire. He’d sleep 12 hours one night and then barely at all the next. He stopped seeing friends and found himself drinking heavily. Around the same time, he started having major mood swings several times a month.
“I mean, you can be in the deep, deep depths of depression and think, ‘I’m going to close my eyes and I’m not going to wake up in the morning,’ ” he said. “And all of a sudden, you wake up on the sunrises, the birds are singing, and you’re like, ‘Wow, where have I been?’ ”
Eventually, Michael went to the doctor and was diagnosed with rapid cycling bipolar disorder. It’s defined as experiencing four or more episodes of mania or depression in a single year. By contrast, people with bipolar that is not rapid cycling experience an average of two mood shifts per year.
Rapid cycling bipolar disorder is also associated with higher suicide risk, poorer long-term prognosis, and greater resistance to treatment than non-rapid cycling bipolar disorder. (Some researchers don’t consider it a different type of bipolar, but simply a more severe presentation of the illness.)
Michael’s doctor put him on a string of medications, but none of them seemed to work, and he continued swinging from depression to mania just about every week.
“It doesn’t give you any security or base,” he said. “If you think about the biblical story about building your house on sand — it’s like waking up in the morning and building your house on a rock and you think, ‘I’m secure, I’m ready to go.’ And then by noon that day, everything’s quaking around you, and you find out you’re on sand and everything’s falling apart.”
In his desperation, Michael did the last thing he could think of to figure out what was causing his mood switches: He started doing his own research.
“I started to write down data,” he said. “And all I could think of to do is write down when you go to sleep, when you got up? What’d you eat that day? What was the weather like? What was the moon phase? You know all these things, I thought, something’s going on here.”
Michael continued to spiral downward, and in 2004 he checked into a local hospital with suicidal thoughts and visual and auditory hallucinations.
David Avery was the inpatient psychiatrist assigned to care for him.
“And he presented to me … a sleep log that he had kept over the previous four months,” Avery said. “And he kept a sleep log because he noted that his sleep-wake cycle had a very clear pattern to it associated with his mood swings.”
Michael’s sleep schedule would shift on almost a weekly basis, between long sleeps associated with depression and short sleeps associated with mania.
“His sleep was such that every 15 days he would sleep for only about zero to two hours per night,” Avery said. “And during that, he would become manic. And seven days after that short sleep, he would sleep for a long period of time and be depressed.”
It was so regular, so weirdly consistent, that Avery figured there had to be some external force causing Michael’s sleep pattern to oscillate from week to week. He had his suspicions about what that force might be — but it wasn’t until more than a decade later that those suspicions would be confirmed.
Laying the groundwork
Psychiatrist Tom Wehr’s fascination with rapid cycling bipolar disorder started in the 1970s, when he was a young researcher at the National Institute of Mental Health.
“The patients there — some of them had a very remarkable form of bipolar illness in which they would be depressed for, you know, three, four weeks and then manic for three or four weeks, and then depressed again,” Wehr said. “And they would just alternate between mania and depression in this fairly rapid manner. And the changes were very dramatic.”
Wehr wanted to know how rapid cycling bipolar disorder — and bipolar in general — functioned: What caused the mood shifts, and how it functioned in the body and mind. So he decided to take it on as a research project.
“I thought, if one were ever going to understand mood disorders and bipolar disorder in particular, this is exactly the type of course of the illness that you could study over time continuously and measure things quantitatively, because the changes were so dramatic and they were so predictable,” he said. “So that kind of led me to focus on that particular pattern of the illness.”
Wehr took what he describes as a “19th-century naturalist” approach. Instead of pursuing a specific hypothesis, he and his research team decided to take a broad approach. They enlisted about 20 research subjects and tracked as many biological variables as possible — their diet, their sleep patterns, their hormones, motor activity, bloodwork … and of course, their moods.
“I was convinced that there were some very powerful biological forces at play,” Wehr said. “But we knew so little about it, that I thought we should just keep an open mind.”
His hope was that by tracking everything, they’d be able to see whether a pattern was emerging — the thing that was causing people to switch from mania to depression and back.
It was a slow-going process. But in the 1990s, Wehr made a discovery that started getting to the root of mood switches in people with rapid cycling bipolar disorder.
“It turns out, if someone is depressed in this way and you simply ask them to stay awake all night, about two-thirds of the time, that person will become manic,” Wehr said.
In 1996, he and his colleagues released a study showing that lack of sleep was the No. 1 predictor of rapid cycling subjects’ switch from depression to mania.
Scientists had already known that sleep disturbances are a part of bipolar disorder. Depression is associated with excessive sleep, and mania with hardly any. But what Wehr’s study showed was that sleep wasn’t just a symptom of the illness — it was part of the cause.
The discovery was part of a wider body of knowledge that Wehr had spent the last few decades expanding. Through his work at the National Institute of Mental Health, he had made a name for himself as an expert on sleep, mood disorders, and circadian rhythms.
He and another colleague were the first to describe seasonal affective disorder, and to propose light therapy as a treatment.
All that work helped lay the groundwork for Wehr’s bigger questions about bipolar disorder: What was it that caused rapid cyclers to switch between mania and depression? And what could be done to stop it?
The hand that winds the clock
It took decades of research before Wehr was finally able to develop a hypothesis. And that hypothesis hinged on a crucial observation about the rapid cycling subjects he’d been observing.
“These people were shifting on a very regular basis between mania and depression,” he said. “Some of these cycles seem so precise in their timing and duration that it was hard to understand how a purely biological process could produce that degree of precision. “
Wehr started wondering if there could be an external signal that was affecting people’s mood cycles.
He’d noticed that, for some of his patients, sleep disturbances seemed to follow regular patterns. Having established that sleep changes were part of what caused switches between depression and mania, Wehr turned his attention to possibilities that affect circadian rhythms.
“It seemed like a biological system could not operate in such a precise manner — there had to be some clock-like external source,” he said. “And, you know, there’s always been a popular belief that the moon is somehow affecting human behavior.”
A belief, but one that’s been mostly debunked by science. And in fact, when Wehr looked at the data, he couldn’t find a connection — the mood and sleep cycles did not seem to correspond with the moon’s phases.
So Wehr kept on thinking. When he had time, he’d take out the data from the 20 or so patients he’d started following back in the ’70s. He’d pore over the cycling patterns, and then put them away again.
This went on for almost three decades before Wehr finally had a break in the case. He thought: What if the mood cycling did have something to do with the moon’s phases … just not the ones related to illumination?
So he began looking into tidal cycles, which are defined by the moon’s gravitational pull on the earth’s oceans. That pull leads to twice-daily high tides and twice-daily low tides. It also governs the spring-neap tidal cycle, which refers to the bimonthly high and low tides (referred to as spring and neap tides, respectively).
What Wehr realized after studying the spring-neap tidal cycle was that, for many of his patients, mania seemed to correspond with the bimonthly spring tides, and depression with the bimonthly neap tides.
The mechanism, he hypothesized, was sleep.
The sun, the moon, and our internal clocks
Sleep cycles are controlled by a little gadget in the brain called the circadian pacemaker, better known as our internal clock. It comprises two networks: one that tracks when we go to sleep, and another that tracks when we wake up.
The sun — or more specifically, light — sets our circadian clocks. Since the solar day is 24 hours, that ideally leads to people sleeping at night and waking up when it gets light.
But in people with bipolar illness, Wehr said, that clock often runs a little slow.
“People who have that problem, they tend to stay up late and have trouble getting up in the morning,” he said. “In other words, they’re kind of predisposed to break away because they’re lagging behind the solar cycle to begin with.”
According to Wehr, once someone’s internal clock falls out of sync with the sun, it becomes vulnerable to syncing up with something else — namely, the gravitational pull of the moon.
But the clock doesn’t uncouple from the sun completely.
“It’s sort of a battle between the sun and the moon for control of this system,” Wehr said.
In his own patients, Wehr theorized that the circadian network responsible for falling asleep stuck with the sun, while the other, which controls wake time, became hijacked by the moon.
That presents a problem because, while the solar day is 24 hours, the lunar day is 24 hours, 50 minutes. In practical terms, that means that the moon rises 50 minutes later every day — and that its peak gravitational force also happens 50 minutes later every day. It also meant that Wehr’s study subjects were waking up later and later every day.
“And so, as those two networks were going in and out of phase with each other,” he said, “they would stretch the duration of sleep longer and longer and longer, and then at a certain point shift and caused sleep to be very short again, as the lunar controlled network would go on around the clock.”
That’s when the spring tides would hit, causing the subjects to stay up all night and shift into mania.
There was just one problem with Wehr’s theory: The relationship was not exact. It wasn’t like every spring tide, every patient would go into mania. At least, none that Wehr had heard of. Until, that is, he heard about Michael.
When Michael checked himself into the hospital in 2004, Tom Wehr was still 13 years away from publishing his first paper on rapid cycling bipolar and the moon. But his psychiatrist, David Avery, had read a different paper by Wehr — a case study that demonstrated success treating rapid cycling bipolar disorder using the same kind of light therapy that’s usually reserved for seasonal affective disorder.
Avery had also read Wehr’s other study identifying sleep disturbances as a cause, rather than a symptom, of bipolar mood cycling. That knowledge, coupled with the drastic disturbances recorded in Michael’s logs, led Avery to target Michael’s sleep schedule above all else.
He used a light box to help re-establish Michael’s circadian rhythm, and kept him on a strict sleep-wake schedule — something Michael says he kept up even after leaving the hospital.
“Whatever he told me to do, I would do because at that time I was so desperate,” Michael said. “He was actually my last hope. I honestly felt like this is going to kill me.”
After 11 days, Michael was discharged, with strict orders to maintain his regular sleeping schedule.
“It was painful,” Michael said. “The first month, I would lay there for hours — and I never laid down for hours like that — going, nope. He said don’t move. You need to lay down at this time. Turn the light off at this time. And even if you can’t sleep, lay there, don’t move.”
But after about two months, Michael said, his body adapted. Fifteen years later, he said he continues to keep a strict sleeping schedule — and hasn’t relapsed since.
Connecting the dots
David Avery and Michael lost touch not long after Michael was discharged, but the question of what had caused his unusual sleep disturbances remained lodged in Avery’s brain.
He had his own theory. Back when Michael was first admitted, Avery was astounded to see that Michael’s sleep cycling, and therefore his mood cycling, was in sync with the lunar tidal cycle. But this was only one case — not much of a basis to publish on — so Avery stuck the data back in a file drawer. That’s where it stayed until 2017, when Wehr published a paper showing the same phenomenon in 17 of his own patients.
For Avery, it was a sign that he’d been on the right track. He contacted Wehr, and together the two men did a deep-dive analysis of Michael’s sleep patterns. Avery had been right — unlike Wehr’s patients, Michael was almost perfectly in sync with the moon’s tidal cycle. Avery and Wehr published their results in 2018.
The collaboration was validating for both men, who’d independently suspected the same phenomenon.
“I think he’s brilliant,” Avery said, referring to Wehr. “I had the advantage of seeing this cycle that was very clearly synchronized with the spring-neap cycle. In his rapid cycling bipolar patients, the cycling was much, much more complex.”
Part of Wehr’s theory is that the moon’s gravitational cycle doesn’t affect everyone the same way. Some people only go into mania every second or third spring tide, he said; they’re all on their own individual schedules with respect to how the moon affects their sleep.
Wehr’s brilliance, Avery said, was his ability to pick out that connection — to figure out that there was a greater pattern underlying each individual’s idiosyncratic response.
Avery and Wehr published a paper together in 2018 using Michael as their case study. It marks an undeniable departure from scientific consensus regarding the moon’s effects on human moods and behavior. But Wehr said he’s unconcerned about skepticism.
“You know, everything that’s new tends to be greeted skeptically, and rightly so,” he said. “In order to discover something new, you have to kind of indulge your imagination and, and fantasy, and kind of think outside the box.”
It’s something that Wehr’s become known for over his decades-long career.
“I think that many of the ideas that he has brought forth into the scientific realm have been very creative and innovative, and on several levels, rather provocative, and caused us all to think in different ways,” Melvin McInnis, a professor of psychiatry and director of the Heinz C. Prechter Bipolar Research Program at the University of Michigan, said. “So it’s very exciting that this is now in the literature.”
While McInnis and his colleague Helen Burgess — who co-directs the University of Michigan’s sleep and circadian research lab — say that the study will need to be replicated, its findings reinforce what sleep and bipolar researchers have been saying for awhile.
“Really, it’s the light/dark cycle that is the strongest influence on the circadian system, which also regulates mood and sleep,” Burgess said. “And once that signal is really solidified and sort of magnified, the paper actually showed the effects of the lunar cycle completely disappeared.”
McInnis also praised Michael for his staunch data-gathering and discipline in sticking to his prescribed sleep schedule.
“If you’ve ever tried to keep a diary of anything, it is notoriously difficult,” McInnis said. “They were fortunate in engaging that individual in research and being able to learn as much as they possibly can from that person.”
Michael is more modest about the part he played in Wehr and Avery’s study.
“I’m just a guy who wrote things down trying to save my life,” he said. “If somebody is out there going through what I did, all I can say is, do not give up. Be disciplined and do not ever give up. There is hope.”
Editor’s notes: This article has been updated to correct the sequence of events leading to David Avery and Tom Wehr’s collaboration.
The audio version of this story mistakenly states where Wehr and Avery jointly published their shared study. Their 2018 paper appeared in the journal Bipolar Disorders.