Home / Sleep Science Hub / Circadian Timing

The Hive Mind · Issue 020 · 16 min read

In 1938, a scientist descended into Mammoth Cave to spend 32 days underground. What he discovered changed everything we know about how women 50+ lose their natural sleep rhythm — and how to get it back.

In the summer of 1938, a University of Chicago physiologist named Nathaniel Kleitman descended into Mammoth Cave in Kentucky carrying a cot, a stack of research notebooks, and enough food to last thirty-two days. He brought one assistant — a graduate student named Bruce Richardson — and together they set up a small camp roughly a hundred feet below the surface, in a chamber where the air stayed a constant 54 degrees Fahrenheit and the darkness was so complete you could hold your hand against your face and see nothing at all.

Kleitman was forty-three years old and already the most serious student of sleep science alive. He had spent more than a decade cataloguing the biology of sleep, and he had grown increasingly convinced that the field was missing something foundational. The question he had come underground to answer was deceptively simple: without the sun, without the ordinary markers of a human day, would the body still know what time it was?

He had designed the experiment carefully. The two men would live on an artificial twenty-eight-hour day. They would sleep when the clock said to sleep and wake when it said to wake. Their meals, their light sources, their activity — everything would follow the new schedule. Kleitman wanted to know whether the human circadian rhythm was truly internal, a clock ticking independently inside the body, or whether it was just a learned response to the world outside.

What happened instead was not what he planned.

Bruce Richardson, twenty-one years old and adaptable in the way young people often are, shifted to the new schedule within a few days. His body temperature, his alertness, his appetite — all of it reorganized around the twenty-eight-hour artificial cycle. Kleitman, nearly twice his age, never made the adjustment. His body clung to something close to the twenty-four-hour rhythm he had carried for four decades, even in the featureless dark. He was tired when the artificial clock said he should be awake. He lay on his cot staring at the ceiling of the cave — and later wrote in his notes that he could not account for what his body was responding to, since there was nothing to respond to. No sunrise. No meal smells drifting in from somewhere. No other person moving around on a schedule that matched his own.

That last detail. He would come back to it.

The clock that doesn't reset itself

Kleitman came out of the cave in August of 1938 with data and a deepened puzzle. He had confirmed that the body does carry something like an internal clock — it doesn't simply go dark without external cues. But he had also discovered that the clock drifts. Left entirely to its own devices, without the usual signals from the environment, it does not stay exactly at twenty-four hours. It wanders. Richardson's clock had been flexible enough to wander toward twenty-eight. Kleitman's was rigid enough that it couldn't.

The implication was uncomfortable and important: the internal clock is real, but it is not self-sufficient. It requires input. It requires what would eventually be called, by a German chronobiologist named Jürgen Aschoff working in the 1950s and 1960s, a Zeitgeber.

The word is German. Zeit means time. Geber means giver. A Zeitgeber is literally a time-giver — any external cue that the brain uses to synchronize its internal clock to the rhythms of the actual world.

Light is the most powerful Zeitgeber. That much was already suspected in Kleitman's era and has since been confirmed in exacting detail. The retinohypothalamic tract — a pathway discovered in the 1970s and characterized more fully through the 1990s — runs from a specific class of retinal cells directly to the suprachiasmatic nucleus, a pair of tiny structures in the hypothalamus that contain roughly twenty thousand neurons and function as the body's master pacemaker. Morning light hits those retinal cells, the signal travels the tract, and the suprachiasmatic nucleus updates its calibration. It sets itself to now.

But Aschoff, who spent decades running isolation experiments in a specially built underground bunker near Munich, found something that Kleitman's cave had hinted at but not fully explained. When he stripped away light — when his subjects lived in constant dim illumination with no day-night variation — their clocks did not simply stop responding to the world. They found other cues. Temperature. Meal timing. Sound. And, with a consistency that surprised him, social contact.

When people in the bunker interacted with other people — even through an intercom, even through a note slid under a door — their rhythms showed a faint but measurable pull toward synchrony. Not a complete reset. Not the sharp recalibration that sunlight produces. Something subtler. A gentle tug.

Aschoff published his observations across a series of papers through the 1960s. He began calling social cues secondary Zeitgebers. Light was primary. But the social world, it turned out, was also in the business of setting clocks.

What the body learns from other people

The mechanism behind social synchrony took decades to work out, and the working-out is still not entirely complete. But the broad shape of it emerged from several converging lines of research in the 1980s and 1990s.

The first line came from mood disorder research. A psychiatrist at the University of Pittsburgh named Ellen Frank noticed in the mid-1980s that many of her patients with depression and bipolar disorder shared a particular history: their episodes had been triggered not by any dramatic emotional event, but by a disruption to their daily schedules. A job that changed shifts. A partner who moved out. A child who grew up and left home. The loss of a friend who had, without either person realizing it, been providing the structural rhythm of the day — the phone call at eight in the morning, the shared lunch, the walk in the evening.

Frank developed a therapeutic approach she called Interpersonal and Social Rhythm Therapy, published formally in 1994, based on a simple premise: that stabilizing the social rhythms of people with mood disorders — regularizing meal times, wake times, key social interactions — stabilized their circadian rhythms, which stabilized their mood. The therapy worked. Not because it addressed trauma or cognition directly, but because it restored the time-giving function that human social structure provides.

The second line came from animal research. Studies in mice and rats through the 1990s — several coming out of the lab of circadian biologist Joseph Takahashi at Northwestern University — showed that isolated animals, kept in constant conditions, had measurably less stable circadian rhythms than animals housed in social groups, even when light conditions were identical for both. The grouped animals weren't just happier. Their clocks were more precise.

The third line was perhaps the most surprising. Researchers studying blind populations — people with no functional light perception at all — found that many of them maintained surprisingly well-organized circadian rhythms despite the absence of the primary Zeitgeber. When researchers looked carefully at what was keeping those rhythms anchored, the most consistent factor was social schedule. Regular meals eaten with other people. Consistent timing of work and conversation and activity. The social world filling in, imperfectly but meaningfully, for the light the eyes could no longer deliver.

By the early 2000s, the picture was clear enough that Till Roenneberg, a chronobiologist at Ludwig Maximilian University of Munich, could summarize it simply: humans are not just biological clocks. We are social clocks. We entrain to each other. We set each other's time.

This connects directly to what we explored in Issue 016, when Karl Kräuchi showed that the body's core temperature must begin its evening descent before sleep can properly begin. That temperature drop is itself partly governed by the circadian clock. And the circadian clock, as we now understand, is set not just by light but by the full texture of the social day — its rhythms, its anchors, its regularities. Change the texture and the temperature curve shifts. The temperature curve shifts and sleep onset moves. The whole system is downstream of those time-giving cues.

The anchor you didn't know you were using

Here is the part that most people find surprising when they first encounter it. The social Zeitgeber does not work only when you are actively engaged with another person. It works through anticipation. Through structure. Through the way a regular social rhythm teaches your brain when the day is, and therefore when the day is over.

Roenneberg's research group published a study in 2007 tracking more than fifty thousand people across Europe and analyzing what predicted the consistency of their circadian timing. Light exposure mattered, as expected. But the consistency of social obligations — particularly the timing of work, shared meals, and regular social contact — was an independent predictor, not reducible to light exposure. People whose social schedules varied widely from day to day had more dispersed, less reliable circadian rhythms than people whose social patterns were stable, even when those people had identical light exposure patterns.

The mechanism, as best as researchers now understand it, works through what is sometimes called the social rhythm as a scaffolding signal. The suprachiasmatic nucleus integrates multiple inputs. Light is the loudest. But the rest of the nervous system is also sending signals up. Cortisol responses to social stressors. Meal-induced metabolic shifts. Temperature changes driven by physical activity. The timing of these signals — if they're consistent — adds a layer of redundant reinforcement to the light signal. The clock gets more input, from more directions. It runs cleaner.

When the social scaffolding disappears, the clock doesn't necessarily stop. But it becomes noisier. More variable. More susceptible to drift.

There is a particular kind of sleeplessness that tends to follow major life transitions — the death of a partner, retirement, children leaving home, a move to a new city. It is usually explained as grief, or stress, or adjustment. Those explanations are not wrong. But there is also something more specific happening. The Zeitgebers that person relied on, without knowing they were relying on them, have been removed. The clock has lost its scaffolding. It is still running, but it is running looser. Bedtime drifts. Wake time drifts. The body temperature curve from Issue 016 shifts and blurs. And the result is exactly the kind of fragmented, unreliable sleep that arrives not from any identifiable pathology but from a subtler impoverishment: the social world no longer knows what time it is, and neither, consequently, does the body.

The drift you recognize at 3am

There is something about the 3am waking that feels different from ordinary insomnia. It is not the inability to fall asleep. It is the inability to stay asleep. You surface in the middle of the night with a peculiar sense of wrongness — not quite anxiety, not quite pain, but a misfiring. An alarm going off in an empty building.

This is the circadian signal behaving erratically. The architecture of sleep — the cycling between deep slow-wave sleep and REM sleep that we explored in Issues 013 and 015 — depends on precise circadian timing. The second half of the night, when those 3am wakings tend to happen, is governed more heavily by circadian signals than the first half, which is driven more by sleep pressure. If the circadian signal is noisy, that second half fractures.

What makes the circadian signal noisy? Light disruption, yes. Temperature dysregulation, yes. But also — and this is the part that rarely makes it into the standard advice — social rhythm disruption. The absence of consistent social anchors across the day means the clock arrives at midnight less precisely set than it should be. It overshoots, or undershoots. It wakes you at 3am with a cortisol pulse that belongs at 6.

This is not a metaphor. Researchers measuring cortisol profiles in people with disrupted social rhythms consistently find that the cortisol awakening response — the sharp rise in cortisol that normally peaks about thirty minutes after waking and helps establish the day — is blunted, shifted, or split into multiple smaller pulses. One of those pulses lands at 3am. That pulse is what wakes you.

The 3am woman reading this on her phone with the brightness turned down is not broken. Her clock is not broken. Her clock is responding, correctly, to the social signal environment it has been receiving. If that environment has changed — because she lives alone now, or because retirement erased her schedule, or because the rhythmic social contact she once had has thinned — the clock has adapted. The adaptation is technically accurate. It is just not what she wants.

Why this matters more after 50

There is a neurological reason that social Zeitgebers become more important, not less, as the years accumulate.

The suprachiasmatic nucleus — the master pacemaker — loses neurons with age. Research published by Erik Herzog at Washington University in St. Louis, building on work from the late 1990s through the 2010s, showed that the electrical coupling between neurons in the suprachiasmatic nucleus weakens with age. The individual cells still oscillate. But the coordination between them becomes less precise. The clock, as a system, produces a less clean signal.

At the same time, the retinal cells that respond to light — specifically the intrinsically photosensitive retinal ganglion cells that carry the primary time-setting signal to the suprachiasmatic nucleus — decrease in sensitivity with age. The pupil admits less light. The lens yellows slightly, filtering blue wavelengths. The same morning light that once produced a strong, sharp calibration signal now produces a weaker one.

The result is a circadian system that is running on reduced input from both its primary signal source and its internal signal-processing capacity. What fills in the gap matters more than it used to.

A 2016 study by Erin Flynn-Evans at NASA's Ames Research Center, examining shift workers across age groups, found that older workers showed significantly greater circadian disruption in response to schedule irregularity than younger workers — not because they were more vulnerable in some general sense, but because they had less circadian reserve. Their clocks depended more heavily on external scaffolding to maintain precision, and when that scaffolding was inconsistent, the degradation was faster and more pronounced.

This is why so many women in their fifties describe a pattern that sounds like this: sleep was fine until something changed — a retirement, a relationship shift, a move, an adult child leaving — and then suddenly, it wasn't. The change didn't cause a disease. It removed a structural support that the body had always needed but had never consciously recognized. The clock had been quietly relying on the rhythm of a shared life. When the shared life changed, the clock began to drift.

There is also a hormonal layer. The drop in estrogen that accompanies perimenopause and menopause has direct effects on thermoregulation — which is how the body temperature story from Issue 016 intersects with the social rhythm story here. But estrogen also modulates the sensitivity of the suprachiasmatic nucleus to light signals. As estrogen declines, the light-based calibration signal weakens further, making the secondary Zeitgebers — social rhythm, meal timing, temperature cues — carry even more of the load. The scaffolding becomes more important precisely as the body's other calibration tools become less reliable.

This is not a counsel of despair. It is a map. And like any map, it shows not just where you are but where the path runs.

· · ·

Until next issue

Next: Next issue: the science of why your gut keeps you awake — and the specific gut-brain signal that has to quiet before sleep can begin.

---

About Beezy Beez. Beezy Beez crafts botanical extract honey for people navigating sleep changes after 50. The Hive Mind is the brand's editorial letter on the science and history of rest.

← Back to the Sleep Science Hub